1: \documentstyle[aps,epsfig]{revtex} 

2: %\documentstyle[12pt]{article}

3: \begin{document} 

4: \draft

5: \preprint{gr-qc/yymmxxx} 

6: \title{\large {\bf Logarithmic correction to the Bekenstein-Hawking entropy}}

7: \author{Romesh K. Kaul and Parthasarathi 

8: Majumdar\footnote{email: kaul, partha@imsc.ernet.in}}

9: \address{\it The Institute of Mathematical Sciences, Chennai 600 113, India.}

10: \maketitle

11: \begin{abstract}

12: The exact formula derived by us earlier for the entropy of a four 

13: dimensional non-rotating black hole within the quantum geometry formulation 

14: of the event horizon in terms of boundary states of a three dimensional 

15: Chern-Simons theory, is reexamined for large horizon areas. In addition to

16: the 

17: {\it semiclassical} Bekenstein-Hawking contribution proportional to the 

18: area obtained earlier, we find a contribution proportional to the logarithm of

19: the area together with subleading corrections that constitute a series in 

20: inverse powers of the area.  

21: \end{abstract}

22: 

23: The derivation of the Bekenstein-Hawking (BH) area law for black hole

24: entropy from the quantum geometry

25: approach \cite{ash} (and also earlier from string theory \cite{stro} for

26: some

27: special cases), has lead to a resurgence of interest in the quantum aspects of

28: black hole physics in recent times.  However, the major activity has remained

29: focussed on {\it confirming} the area law for large black holes, which, as

30: is well-known, was obtained

31: originally on the basis of arguments of a semiclassical nature. The question

32: arises as to whether any essential feature of the bona fide quantum aspect of

33: gravity, beyond the domain of the semiclassical approximation, has been captured

34: in these assays. Indeed, as has been most eloquently demonstrated by Carlip

35: \cite{car}, a derivation of the area law alone seems to be possible on the basis

36: of some symmetry principle of the (semi)classical theory itself without

37: requiring a detailed knowledge of the actual quantum states associated with a

38: black hole. The result seems to hold for arbitrary number of spatial dimensions,

39: so long as a particular set of isometries of the metric is respected. That

40: quantum gravity has a description in terms of spin networks (or for that matter,

41: in terms of string states in a fixed background) appears to be of little

42: consequence in obtaining the area law, although these proposed underlying

43: structures also lead to the same behaviour via alternative routes, in the

44: semiclassical limit of arbitrarily large horizon area. 

45: 

46: Although there is as yet no complete quantum theory of gravitation, one

47: would in general expect key features uncovered so far to lead to

48: modifications of the area law which could not have been anticipated through

49: semiclassical reasoning. Thus, the question as to what is the dominant quantum

50: correction due to these features of quantum gravity becomes one of paramount

51: importance. Already in the string theory literature \cite{wit} examples of

52: leading corrections to the area law, obtained by counting D-brane states

53: describing special supersymmetric extremal black holes (interacting with

54: massless vector supermultiplets) have appeared. This has received strong support

55: recently from semiclassical calculations in $N=2$ supergravity \cite{dew}

56: supplemented by ostensible stringy higher derivative corrections which are

57: incorporated using Wald's general formalism describing black hole entropy as

58: Noether charge \cite{wald}. However, the geometrical interpretation of

59: these corrections remains unclear. Further, there are subtleties

60: associated with direct application of Wald's formalism which assumes a

61: {\it non-degenerate} bifurcate Killing horizon, to the case of

62: extremal black holes which have degenerate horizons. Moreover, the string

63: results do not pertain to generic (i.e., non-extremal) black holes of

64: Einstein's general relativity, and are constrained by the unphysical

65: requirement of unbroken spacetime supersymmetry. 

66: 

67: In this paper, we consider the corrections to the semiclassical area law

68: of generic four dimensional non-rotating black holes, due to key aspects

69: of {\it non-perturbative} quantum gravity (or quantum geometry) 

70: formulated by Ashtekar and collaborators \cite{ash2}. In \cite{ash},

71: appropriate boundary conditions are imposed on dynamical variables at the

72: event horizon considered as an inner boundary. These boundary conditions

73: require that the Einstein-Hilbert action be supplemented by boundary terms

74: describing a three dimensional $SU(2)$ Chern-Simons theory living on a

75: finite `patch' of the horizon with a spherical boundary, punctured by

76: links of the spin network bulk states describing the quantum spacetime

77: geometry interpolating between asymptopia and the horizon. On this two

78: dimensional boundary there exists an $SU(2)$ Wess Zumino model whose

79: conformal blocks describe the Hilbert space of the Chern-Simons theory

80: modelling the horizon. An exact formula for the number of these conformal

81: blocks has been obtained by us \cite{km} two years ago, for arbitrary level

82: $k$ and number of punctures $p$. It has been shown that in the limit of large

83: horizon area given by arbitrarily large $k$ and $p$, the logarithm of

84: this number duly yields the area law. Here we go one step further, and

85: calculate the dominant sub-leading contribution, as a function of the

86: classical horizon area, or what is equivalent, as a function of the BH

87: entropy itself. 

88: 

89: On purely dimensional grounds, one would expect the entropy to have an

90: expansion, for large classical horizon area, in inverse powers of area so

91: that the BH term is the leading one,

92: \begin{equation}

93: S_{bh}~~=~~S_{BH}~+~\sum_{n=0}^{\infty} C_n~A_H^{-n}~~\label{gen}

94: \end{equation}

95: where, $A_H$ is the classical horizon area and $C_n$ are coefficients

96: which are independent of the horizon area but dependent on the Planck

97: length (Newton constant). Here the Barbero-Immirzi parameter \cite{barb}  

98: has been `fitted' to the value which fixes the normalization of the BH term

99: to the standard one. However, in principle, one could expect an additional

100: term proportional to $ln~A_H$ as the leading quantum correction to the

101: semiclassical $S_{BH}$. Such a term is expected on general grounds

102: pertaining to breakdown of na\"ive dimensional analysis due to quantum

103: fluctuations, as is common in quantum field theories in flat spacetime

104: and also in quantum theories of critical phenomena. We show, in what

105: follows, that such a logarithmic correction to the semiclassical area law

106: does indeed arise from the formula derived earlier \cite{km} and derive

107: its coefficient. 

108: 

109: We first briefly recapitulate the derivation \cite{km} of the general

110: formula for the number of conformal blocks of the $SU(2)_k$ Wess Zumino

111: model on a punctured 2-sphere appropriate to the black hole situation. This

112: number can be computed in terms of the so-called fusion matrices

113: $N_{ij}^{~~r}$ \cite{dms} 

114: \begin{equation} 

115: N_{\cal P}~=~~\sum_{\{r_i\}}~N_{j_1 j_2}^{~~~~r_1}~ N_{r_1 j_3}^{~~~~r_2}~

116: N_{r_2 j_4}^{~~~~r_3}~\dots \dots~ N_{r_{p-2} j_{p-1}}^{~~~~~~~~j_p} ~

117: \label{fun} \end{equation} 

118: Diagrammatically, this can be represented as shown in fig. 1 below. 

119: \begin{figure}[htb] \begin{center}

120: \mbox{\epsfig{file=bhefig1.eps,width=10truecm,angle=0}} 

121: \caption{Diagrammatic representation of composition of spins $j_i$ for

122: $SU(2)_k$} 

123: \end{center}

124: \end{figure} 

125: Here, each matrix element $N_{ij}^{~~r}$ is $1 ~or~ 0$,

126: depending on whether the primary field $[\phi_r]$ is allowed or not in the

127: conformal field theory fusion algebra for the primary fields $[\phi_i]$

128: and $[\phi_j] $ ~~($i,j,r~ =~ 0, 1/2, 1, ....k/2$): 

129: \begin{equation}

130: [\phi_i] ~ \otimes~ [\phi_j]~=~~\sum_r~N_{ij}^{~~r} [\phi_r]~ . 

131: \label{fusal}

132: \end{equation}

133: Eq. (\ref {fun}) gives the number of conformal blocks with spins $j_1,

134: j_2, \dots, j_p$ on $p$ external lines and spins $r_1, r_2, \dots,

135: r_{p-2}$ on the internal lines. 

136: 

137: We then use the Verlinde formula \cite{dms} to obtain

138: \begin{equation}

139: N_{ij}^{~~r}~=~\sum_s~{{S_{is} S_{js} S_s^{\dagger r }} \over S_{0s}}~,

140: \label{verl} 

141: \end{equation} 

142: where, the unitary matrix $S_{ij}$ diagonalizes the fusion matrix. Upon

143: using the unitarity of the $S$-matrix, the algebra (\ref{fun}) reduces to 

144: \begin{equation} 

145: N_{\cal P}~=~ \sum_{r=0}^{k/2}~{{S_{j_1~r} S_{j_2~r} \dots S_{j_p~r}}

146: \over (S_{0r})^{p-2}}~. 

147: \label{red} \end{equation} 

148: Now, the matrix elements of $S_{ij}$ are known for the case under

149: consideration ($SU(2)_k$ Wess-Zumino model); they are given by

150: \begin{equation} 

151: S_{ij}~=~\sqrt{\frac2{k+2}}~sin \left({{(2i+1)(2j+1) \pi} \over k+2}

152: \right )~, \label{smatr}

153: \end{equation} 

154: where, $i,~j$ are the spin labels, $i,~j ~=~ 0, 1/2, 1, .... k/2$. Using

155: this $S$-matrix, the number of conformal blocks for the set of punctures

156: ${\cal P}$ is given by

157: \begin{equation} 

158: N_{\cal P}~=~{2 \over {k+2}}~\sum_{r=0}^{ k/2}~{

159: {\prod_{l=1}^p sin \left( {{(2j_l+1)(2r+1) \pi}\over k+2} \right) } \over

160: {\left[ sin \left( {(2r+1)  \pi \over k+2} \right)\right]^{p-2} }} ~. 

161: \label{enpi} \end{equation} 

162: Eq. (\ref{enpi}) thus gives the dimensionality of the $SU(2)$ Chern-Simons

163: states corresponding to a three-fold bounded by a two-sphere punctured at

164: $p$ points. The black hole microstates are counted by summing $N_{\cal P}$

165: over all sets of punctures ${\cal P}$, $N_{bh}=\sum_{\{\cal P\}} N_{\cal

166: P}$. Then, the entropy of the black hole is given by $S_{bh}~=~\log

167: N_{bh}$. 

168: 

169: We are however interested only in the leading correction to the semiclassical

170: entropy which ensues in the limit of arbitrarily large $A_H$. To this end,

171: recall that the eigenvalues of the area operator \cite{ash2} are given by

172: \begin{equation}

173: A_H~=~8\pi \beta ~l_{Pl}^2~\sum_{l=1}^p~[j_l(j_l+1)]^{\frac12}~,

174: \label{area} \end{equation}

175: where, $l_{Pl}$ is the Planck length, $j_l$ is the spin on the $l$th

176: puncture on the 2-sphere and $\beta$ is the Barbero-Immirzi parameter

177: \cite{barb}. Clearly, the large area limit corresponds to the limits $k

178: ~\rightarrow~\infty~,~p~\rightarrow~\infty$. Now, from eq. (\ref{area}), it

179: follows that the number of punctures $p$ is largest for a given $A_H$

180: provided {\it all} spins $j_l~=~\frac12$. Thus, for a fixed classical

181: horizon area, we obtain the largest number of punctures $p_0$ as

182: \begin{equation}

183: p_0~=~{A_H \over 4 l_{Pl}^2}~{\beta_0 \over \beta}~, \label{pmax}

184: \end{equation}

185: where, $\beta_0=1/\pi \sqrt{3}$. In this approximation, the set of punctures

186: ${\cal P}_0$ with all spins equal to one-half dominates over all other

187: sets, so that the black hole entropy is simply given by 

188: \begin{equation}

189: S_{bh}~~=~~ln~N_{{\cal P}_0}~~, \label{ent}

190: \end{equation}

191: with $N_{{\cal P}_0}$ being given by eq. (\ref{enpi}) with $j_l =

192: 1/2$. 

193: 

194: Observe that $N_{{\cal P}_0}$ can now be written as 

195: \begin{equation}

196: N_{{\cal P}_0}~=~{2^{p_0+2} \over k+2}~\left[

197: F(k,p_o)~-~F(k,p_0+2)

198: \right]~\label{hlf} \end{equation}

199: where, 

200: \begin{equation}

201: F(k,p)~=~\sum^{[\frac12(k+1)]}_{\nu=1}~cos ^p \left({\nu \pi \over k+2}

202: \right)~. \label{eff} \end{equation}

203: The sum over $\nu$ in eq. (\ref{eff}) can be approximated by an integral

204: in the limit $k~\rightarrow~\infty~,~p_0~\rightarrow~\infty$, with

205: appropriate care being taken to restrict the domain of integration; one

206: obtains

207: \begin{equation}

208: F(k,p_0)~\approx~\left({k+2 \over \pi}

209: \right)~\int_0^{\pi/2}~dx~cos^{p_0} x~, \label{inte} \end{equation}

210: so that,

211: \begin{equation}

212: N_{{\cal P}_0}~\approx~{2^{p_0+2} \over \pi (p_0 +2)}~B~({p_0+1

213: \over 2}~,~\frac12)~, \label{nap} 

214: \end{equation}

215: where, $B(x,y)$ is the standard $B$-function \cite{whit}. Using well-known

216: properties of this function, it is straightforward to show that

217: \begin{eqnarray}

218: ln~N_{{\cal P}_0}&~=~&

219: p_0~ln2~-~\frac32~ln~p_0~-~ln~(2\pi)~\nonumber \\

220: &-&~\frac52~p_0^{-1}~+~O(p_0^{-2})~~. \label{lnn}

221: \end{eqnarray}

222: Substituting for $p_0$ as a function of $A_H$ from eq. (\ref{pmax}) and

223: setting the Barbero-Immirzi parameter $\beta$ to the `universal' value

224: $ \beta_0~ln2 $ \cite{ash}, one obtains our main result

225: \begin{equation}

226: S_{bh}~=~S_{BH}~-~{3\over 2}~ln\left( S_{BH} \over ln2

227: \right)~+~const.~+~\cdots~, \label{main}

228: \end{equation}

229: where, $S_{BH}=A_H/4 l_{Pl}^2$, and the ellipses denote corrections in

230: inverse powers of $A_H$ or $S_{BH}$. 

231: 

232: Admittedly, the above calculation is restricted to the leading correction

233: to the semiclassical approximation. It has been done for a fixed large

234: $A_H$ by taking the spins on all the punctures to be 1/2 so that we have

235: the largest number of punctures. But it is not difficult to argue that

236: the coefficient of the $ln A_H$ term is robust in that inclusion of spin

237: values higher than 1/2 do not affect it, although the constant term and

238: the coefficients of sub-leading corrections with powers of $O(A_H^{-1})$

239: might get affected. The same appears to be true for values of

240: the level $k$ away from the asymptotic value which we have assigned it

241: above: the coefficient of the $ln A_H$ is once again unaffected. Thus, the

242: leading logarithmic correction with coefficient -3/2 that we have

243: discerned for the black hole entropy is in this sense {\it universal}. 

244: Moreover, although we have set $\beta=\beta_0~ln2$ in the above formulae,

245: the coefficient of the $ln A_H$ term is independent of $\beta$, a feature

246: not shared by the semiclassical area law. 

247: 

248: It is therefore clear that the leading correction (and maybe also the

249: subleading ones) to the BH entropy is negative. One way to understand this

250: could be the information-theoretic approach of Bekenstein \cite{bek}: black

251: hole entropy represents lack of information about quantum states which arise

252: in the various ways of gravitational collapse that lead to formation of black

253: holes with the same mass, charge and angular momentum. Thus, the BH entropy

254: is the `maximal' entropy that a black hole can have; incorporation of

255: leading quantum effects reduces the entropy. The logarithmic nature of

256: the leading correction points to a possible existence of what might be

257: called a `non-perturbative fixed point'. That this happens in the physical

258: world of four dimensions is perhaps not without interest. 

259: 

260: Recently, the zeroth and first law of black hole mechanics have been

261: derived for situations with radiation present in the vicinity

262: of the horizon, using the notion of the isolated horizon \cite{ash3}. Our

263: conclusions above for the case of non-rotating black holes hold for such

264: generalizations \cite{ash4} as well. Note however that while, 

265: the foregoing analysis involves $SU(2)_k$ Chern Simons theory, for large 

266: $k$ this reduces to a specific $U(1)$ theory presumably related to 

267: the `gauge fixed' classical theory discussed in \cite{ash3}. The charge 

268: spectrum of this $U(1)$ theory is discrete and bounded from 

269: above by $k$. The $SU(2)$ origin of the theory thus provides a natural 

270: `regularization' for calculation of the number of conformal blocks. 

271: 

272: \noindent {\it Note Added:} After the first version of this paper appeared

273: in the Archives, it has been brought to our attention that corrections to

274: the area law in the form of logarithm of horizon area have been obtained

275: earlier \cite{mann} for extremal Reissner-Nordstrom and dilatonic black

276: holes. These corrections are due to quantum scalar fields propagating in

277: fixed classical backgrounds appropriate to these black holes. The

278: coefficient of the $ln A_H$ term that appears in ref. \cite{mann} is

279: different from ours. This is only expected, since in contrast to ref.

280: \cite{mann}, our corrections originate from non-perturbative quantum

281: fluctuations of spacetime geometry (for generic non-rotating black holes),

282: {\it in the absence of matter fields.} Thus, this correction is {\it

283: finite} and independent of any arbitrary `renormalization scale'

284: associated with divergences due to quantum matter fluctuations in a fixed 

285: classical background. 

286: 

287: We thank Prof. A. Ashtekar for many illuminating discussions and Prof. R.

288: Mann for useful correspondence.

289: 

290: \begin{references}

291: \bibitem{ash} A. Ashtekar, J. Baez, A. Corichi and K. Krasnov, {\it Phys.

292: Rev. Lett.} {\bf 80}, 904 (1998).

293: \bibitem{stro} A. Strominger and C. Vafa, {\it Phys. Lett.} {\bf B379}, 99

294: (1996).

295: \bibitem{car} S. Carlip, {\it Class. Quant. Grav.} {\bf 16}, 3327 (1999).

296: \bibitem{wit} J. Maldacena, A. Strominger and E. Witten, {\it Jour. High

297: Energy Phys.} {\bf 12}, 2 (1997).

298: \bibitem{dew} B. de Wit, Modifications of the area law and $N=2$

299: supersymmetric black holes, hep-th/9906095 and references therein. 

300: \bibitem{wald} R. Wald, {\it Phys. Rev.} {\bf D48}, 3427 (1993); T.

301: Jacobson, G. Kang and R. Myers, {\it Phys. Rev.} {\bf D49}, 6587 (1994); V.

302: Iyer and R. Wald, {\it Phys. Rev.} {\bf D50}, 846 (1995).. 

303: \bibitem{ash2} A. Ashtekar, {\it Lectures on Non-perturbative Canonical

304: Gravity}, World Scientific, 1991;  A. Ashtekar and J. Lewandowski in {\it

305: Knots and Quantum Gravity}, ed. J. Baez, Oxford University Press, 1994; {\it

306: Class. Quant. Grav.} {\bf 14}, A55 (1997); J. Baez, {\it Lett. Math. Phys.}

307: {\bf 31}, 213 (1994); C. Rovelli and L. Smolin, {\it Nucl. Phys.} {\bf B331},

308: 80 (1990); {\it Nucl. Phys.} {\bf B442}, 593 (1995). See also references

309: quoted in A. Ashtekar, Interface of General Relativity, Quantum Physics 

310: and Statistical Mechanics: Some Recent Developments, gr-qc/9910101. 

311: \bibitem{km} R. Kaul and P. Majumdar, {\it Phys. Lett.} {\bf B439}, 267

312: (1998). 

313: \bibitem{barb} F. Barbero, {\it Phys. Rev.} {\bf D54}, 1492 (1996); G.

314: Immirzi, {\it Nucl. Phys. Proc. Suppl.} {\bf 57}, 65 (1997). 

315: \bibitem{dms} P. Di Francesco, P. Mathieu and D. Senechal, {\it Conformal

316: Field Theory}, Springer, 1997, p 375.

317: \bibitem{whit} E. Whittaker and G. Watson, {\it Modern Analysis}, Cambridge,

318: 1962. 

319: \bibitem{bek} J. Bekenstein, {\it Phys. Rev.} {\bf D7}, 2333 (1973).

320: \bibitem{ash3} A. Ashtekar, C. Beetle and S. Fairhurst, Mechanics of

321: isolated horizons, gr-qc/9907068, and references therein. 

322: \bibitem{mann} R. Mann and S. Solodukhin, {\it Nucl. Phys.} {\bf 523B}, 

323: 293 (1998) and references therein.

324: \bibitem{ash4} A. Ashtekar, private communication. 

325: \end{references}

326: \end{document}

327: 

328: 

329: