-
The
$ d+1 $ dimensional Einstein-Maxwell theory has an action (in units of$ c = \hbar = 1 $ ) as$ I = \int {\rm d}^{d+1}x \sqrt{-g} \left[ \frac{1}{16 \pi G_{d+1}} \left( R + \frac{d(d-1)}{L^2} \right) - \frac1{g_{\mathrm s}^2} F_{\mu\nu} F^{\mu\nu} \right], $
(1) where L is the curvature radius of the asymptotical
$ {\rm{AdS}}_{d+1} $ spacetime, and$ g_{\mathrm s} $ is the dimensionless coupling constant of the$ U(1) $ gauge field. The dynamical equations$ \begin{aligned}[b] R_{\mu\nu} \!-\! \frac12 g_{\mu\nu} R \!-\! \frac{d(d-1)}{2L^2} g_{\mu\nu} & \!=\! \frac{8 \pi G_{d+1}}{g_{\mathrm s}^2} \left( 4 F_{\mu\lambda} F_{\nu}{}^{\lambda} \!-\! g_{\mu\nu} F_{\alpha\beta} F^{\alpha\beta} \right), \\ \partial_\mu \left( \sqrt{-g} F^{\mu\nu} \right) & = 0, \end{aligned} $
(2) admit the Reissner-Nordström-Anti de Sitter (RN-
$ {\rm{AdS}}_{d+1} $ ) black brane (or the planar black hole) solution [34]$ \begin{aligned}[b] {\rm d}s^2 & = \frac{L^2}{r^2 f(r)} {\rm d}r^2 + \frac{r^2}{L^2} \left( -f(r) {\rm d}t^2 + {\rm d}x_i^2 \right), \\ A & = \mu \left( 1 - \dfrac{r_{\mathrm o}^{d-2}}{r^{d-2}} \right) {\rm d}t, \end{aligned} $
(3) with
$\begin{aligned}[b] f(r) =& 1 - \frac{G_{d+1} L^2 M}{r^d} + \frac{G_{d+1} L^2 Q^2}{r^{2d-2}}, \qquad {} \\ \mu =& \sqrt{\frac{d-1}{2(d-2)}} \frac{g_{\mathrm s} Q}{r_{\mathrm o}^{d-2}},\end{aligned} $
(4) where
$ r_{\mathrm o} $ is the radius of the outer horizon ($ f(r_{\mathrm o}) = 0 $ ),$ \mu $ is the chemical potential with dimension$ [\mu] = \mathrm{length}^{-(d-1)/2} $ , M is the mass, and Q is the charge of the black brane. We may find an explicit expression of$ r_{\mathrm o} $ for$ d = 4 $ from a solution of the cubic equation, which is complicated, but$ r_\mathrm{o} $ has a general expression in the extremal case, i.e.,$ r_* $ in IIB. The condition$ f(r_{\mathrm o}) = 0 $ gives$ M = \dfrac{r_{\mathrm o}^d}{G_{d+1} L^2} + \dfrac{Q^2}{r_{\mathrm o}^{d-2}} $ (which is the Smarr-like relation related to the first law of thermodynamics of the black brane); temperature T and “surface” entropy density s of the black brane are, respectively,$\begin{aligned}[b] T =& \frac{r_{\mathrm o} \, d}{4\pi L^2} \left( 1 - \frac{d-2}{d} \frac{G_{d+1} L^2 Q^2}{r_{\mathrm o}^{2d-2}} \right), \qquad {\rm{}} \\ s = &\frac{1}{4G_{d+1}} \left( \frac{r_{\mathrm o}}{L} \right)^{d-1}. \end{aligned} $
(5) Moreover, the first law of thermodynamics of the dual boundary d-dimensional quantum field is
$ \delta \epsilon = T \delta s + \mu \delta\rho_{\mathrm c}, $
(6) where the “surface” energy and charge densities are, respectively,
$ \begin{aligned}[b]& \epsilon = \frac{d - 1}{16 \pi L^{d-1}} M, \qquad {\rm{}} \qquad \\&\rho_{\mathrm c} = \frac{\sqrt{2(d-1)(d-2)}}{8 \pi g_{\mathrm s} L^{d-1}} Q. \end{aligned} $
(7) Then, it is straightforward to check the Euler relation
$ \left( \frac{d}{d-1} \right) \epsilon = \epsilon + p = T s + \mu \rho_{\mathrm c}, $
(8) where the pressure is
$ p = \dfrac{\epsilon}{d-1} $ , which shows that the dual d-dimensional quantum field theory on the asymptotic boundary is conformal, as expected. -
To make the following analysis convenient, let us introduce the length scale
$ r_*^{2d-2} \equiv \dfrac{d-2}{d} G_{d+1} L^2 Q^2 $ ; then, the temperature can be rewritten as$ T = \frac{r_{\mathrm o} d}{4\pi L^2} \left( 1 - \frac{r_*^{2d-2}}{r_{\mathrm o}^{2d-2}} \right). $
(9) Note that
$ r_* $ may be treated as the “effective” radius of the inner black hole horizon though$ f(r_*) \neq 0 $ in general and$ r_* < r_{\mathrm o} $ . The extremal condition for a degenerate horizon at$ r_\mathrm{o} = r_* $ is$ M = M_0 \equiv \dfrac{2 (d-1)}{d-2} \dfrac{r_*^d}{G_{d+1} L^2} $ . The near extremal limit of the near horizon is obtained by taking the limit$ \varepsilon \to 0 $ of the transformations$ \begin{aligned}[b]& M - M_0 = \frac{d (d-1) r_*^{d-2}}{G_{d+1} L^2} \varepsilon^2 \rho_\mathrm{o}^2, \\& r_{\mathrm o} - r_* = \varepsilon \rho_{\mathrm o}, \quad r - r_{\mathrm o} = \varepsilon (\rho - \rho_0), \quad t = \frac{\tau}{\varepsilon}, \end{aligned}$
(10) where in general
$ \rho_{\mathrm o} $ is finite and$ \rho \in [\rho_0, \infty) $ .Expanding
$ f(r) $ around$ r = r_{\mathrm o} $ , we have$ f(r) \simeq \frac{{d(d - 1)}}{{r_{\rm{o}}^2}}({\rho ^2} - \rho _{\rm{o}}^2){\varepsilon ^2} + O\left( {{\varepsilon ^3}} \right), $
(11) the near horizon geometry is given by
$ \begin{aligned}[b] {\rm d}s^2 & = - \frac{\rho^2 - \rho_{\mathrm o}^2}{\ell^2} {\rm d}\tau^2 + \frac{\ell^2 {\rm d}\rho^2}{\rho^2 - \rho_{\mathrm o}^2} + \frac{r_{\mathrm o}^2}{L^2} {\rm d}x_i^2, \\ A & = \frac{(d-2) \mu}{r_{\mathrm o}} (\rho - \rho_{\mathrm o}) {\rm d}\tau, \end{aligned} $
(12) where
$ \ell^2 \equiv \dfrac{L^2}{d (d - 1)} $ is defined as the square of the curvature radius of the effective AdS2 geometry. The limit$ \rho_{\mathrm o} \to 0 $ yields the extremal limit.The solution in Eq. (12) can also be written in the Poincaré coordinates in terms of
$ \xi = \ell^2/\rho $ , ($ |\xi| \leqslant \xi_{\mathrm o} = $ $ \ell^2/\rho_\mathrm{o} $ ),$ \begin{aligned}[b] {\rm d}s^2 & = \frac{\ell^2}{\xi^2} \left( - \left( 1 - \frac{\xi^2}{\xi_{\mathrm o}^2} \right) {\rm d}\tau^2 + \frac{{\rm d}\xi^2}{1-\dfrac{\xi^2}{\xi_{\mathrm o}^2}} \right) + \frac{r_{\mathrm o}^2}{L^2} {\rm d}x_i^2, \\ A & = \frac{(d-2) \mu \ell^2}{r_{\mathrm o}} \left( \frac{1}{\xi} - \frac{1}{\xi_{\mathrm o}} \right) {\rm d}\tau. \end{aligned} $
(13) The above geometry is a black brane with both local and asymptotical topology
$ {\rm{AdS}}_2 \times {{R}}^{d-1} $ (AdS2 has the$ SL(2,R)_R $ symmetry). The horizons of the new black brane are located at$ \xi = \pm \xi_{\mathrm o} $ , and its temperature is$ T_{ {{n}}} = \dfrac{1}{2 \pi \xi_{\mathrm o}} $ . Note that if we adopt the new coordinates$ z \equiv \xi/\xi_{\mathrm o} $ with$ |z| \leqslant 1 $ and$ \eta = \tau/\xi_{\mathrm o} $ , the metric becomes$ {\rm d}s^2 = \frac{\ell^2}{z^2} \left( - (1 - z^2) {\rm d}\eta^2 + \frac{{\rm d}z^2}{1-z^2} \right) + \frac{r_{\mathrm o}^2}{L^2} {\rm d}x_i^2, $
(14) and the temperature associated with the inverse period of
$ \eta $ is normalized to$ \tilde{T}_{ n} = \dfrac{1}{2\pi} $ . -
The action of a bulk probe charged scalar field
$ \Phi $ with mass m and charge q is$ S = \int {{{\rm d}^{d + 1}}} x\sqrt { - g} \left( { - \frac{1}{2}{D_\alpha^* }{\Phi ^*}{D^\alpha }\Phi - \frac{1}{2}{m^2}{\Phi ^*}\Phi } \right), $
(15) where
$ D_{\alpha} \equiv \nabla_{\alpha} - {\rm i} q A_{\alpha} $ with$ \nabla_\alpha $ being the covariant derivative in curved spacetime. The corresponding Klein-Gordon (KG) equation is$ (\nabla_\alpha - {\rm i} q A_\alpha) (\nabla^\alpha - {\rm i} q A^\alpha) \Phi = m^2 \Phi. $
(16) Moreover, the radial flux of the probe field is
$ {\cal F} = {\rm i}\sqrt { - g} {g^{rr}}(\Phi D_r^*{\Phi ^*} - {\Phi ^*}{D_r}\Phi ). $
(17) In the RN-
$ {\rm{AdS}}_{d+1} $ background (3), assuming$ \Phi(t, \vec{x}, r) = $ $ \phi(r) \mathrm{e}^{-{\rm i} \omega t + {\rm i} \vec{k} \cdot \vec{x}} $ , the KG Eq. (16) has the radial form$ \begin{aligned}[b]& \left(\frac{L}{r}\right)^{d-1} \partial_r \left( \frac{r^{d+1}}{L^{d+1}} f(r) \partial_r \right) \phi(r) \\& + \left(\frac{L^2 (\omega + q A_t)^2}{r^2 f(r)} - m^2 - \frac{L^2}{r^2} \vec{k}^2 \right) \phi(r) = 0. \end{aligned}$
(18) The solutions to Eq. (18) cannot be directly found in terms of special functions in the full spacetime region. In what follows, we solve it in different regions and match these solutions to obtain the full solution.
-
Firstly, we analyze the near horizon, near extreme region (13) and solve the KG Eq. (16) by expanding the scalar field as
$ \Phi(\tau, \vec{x}, \xi) = \phi(\xi) \mathrm{e}^{-{\rm i} w \tau + {\rm i} \vec{k} \cdot \vec{x}}. $
(19) Then, the KG equation reduces to ①
$\begin{aligned}[b] \xi^2 \left( 1 - \frac{\xi^2}{\xi_{\mathrm o}^2} \right) \phi''(\xi) - \frac{2\xi^3}{\xi_{\mathrm o}^2} \phi'(\xi) + \xi^2 \frac{(w + q A_\tau)^2}{1 - \dfrac{\xi^2}{\xi_{\mathrm o}^2}} \phi(\xi) = m_\mathrm{eff}^2 \ell^2 \phi(\xi), \end{aligned} $
(20) where the effective mass square is defined as
$ m_\mathrm{eff}^2 = $ $ m^2 + \dfrac{L^2 \vec{k}^2}{r_{\mathrm o}^2} $ , or the KG equation can be expressed in the z coordinate as$ \begin{aligned}[b] z^2 (1 - z^2) \phi''(z) - 2 z^3 \phi'(z) &+ \frac{z^2}{1 - z^2} \left[ \left( w \xi_{\mathrm o} + q_\mathrm{eff} \ell \frac{1 - z}{z} \right)^2\right. \\&\left.- m_\mathrm{eff}^2 \ell^2 \frac{1 - z^2}{z^2} \right] \phi(z) = 0, \end{aligned} $
(21) where the effective charge of the probe field is
$ q_\mathrm{eff} \equiv $ $ (d-2) \dfrac{\mu \ell}{r_{\mathrm o}} q $ . The singularities of Eq. (21) are located at$ z = 0, z = \pm 1 $ and$ z = \infty $ .To find the solutions, we determine the indices at each singular point. For
$ z \to 0 $ , setting$ \phi(z) \sim z^{\bar{\alpha}} $ , the leading terms in Eq. (21) are$ z^2 \phi''(z) + (q_\mathrm{eff}^2 - m_\mathrm{eff}^2) \ell^2 \phi(z) = 0, $
(22) which gives
$ \begin{aligned}[b] \bar{\alpha} = &\frac12 \pm \frac12 \sqrt{1 + 4 ( m_\mathrm{eff}^2 - q_\mathrm{eff}^2 ) \ell^2} \\\equiv &\frac12 \pm \frac12 \sqrt{1 + 4 \tilde{m}_\mathrm{eff}^2 \ell^2} \equiv \frac12 \pm \nu. \end{aligned} $
(23) For
$ z \to -1 $ , setting$ \phi(z) \sim (1 + z)^{\bar{\beta}} $ , Eq. (21) reduces to$ 2 (1 + z) \phi''(z) + 2 \phi'(z) + \frac{(w \xi_{\mathrm o} - 2 q_\mathrm{eff} \ell)^2}{2 (1 + z)} \phi(z) = 0, $
(24) and the index is
$ \bar{\beta} = \pm {\rm i} \left( \frac{w \xi_{\mathrm o}}{2} - q_\mathrm{eff} \ell \right) = \pm {\rm i} \left( \frac{w}{4 \pi T_{\mathrm n}} - q_\mathrm{eff} \ell \right). $
(25) Finally, for
$ z \to 1 $ , setting$ \phi(z) \sim (1 - z)^{\bar{\gamma}} $ , Eq. (21) reduces to$ 2 (1 - z) \phi''(z) - 2 \phi'(z) + \frac{(w \xi_{\mathrm o})^2}{2 (1 - z)} \phi(z) = 0, $
(26) from which
$ \bar{\gamma} = \pm {\rm i} \frac{w \xi_{\mathrm o}}{2} = \pm {\rm i} \frac{w}{4 \pi T_{\mathrm n}} = \pm {\rm i}\frac{\omega/\varepsilon}{4\pi /(2\pi \xi_{\mathrm o})} = \pm {\rm i}\frac{\omega}{2\varepsilon\rho_{\mathrm o}/\ell^2} = \pm {\rm i}\frac{\omega}{4\pi T} $
(27) is obtained. Further, imposing the ingoing boundary condition at the black brane horizon
$ z = 1 $ requires$ \bar{\gamma} = $ $ -{\rm i} \dfrac{w \xi_{\mathrm o}}{2} = -{\rm i} \dfrac{w}{4\pi T_{\mathrm n}} $ .Also, note that Eq. (21) can be rewritten in a more explicit form as
$ \begin{aligned}[b] \phi''(z) + \left( \frac{1}{z+1} + \frac{1}{z-1} \right) \phi'(z) + \left( \frac{\tilde{m}_\mathrm{eff}^2 \ell^2}{z} + \frac{\dfrac{1}{2} (w \xi_{\mathrm o} - 2 q_\mathrm{eff} \ell)^2}{z+1} + \frac{\dfrac12 w^2 \xi_{\mathrm o}^2}{z-1} \right) \frac{\phi(z)}{z (z+1) (z-1)} = 0, \end{aligned} $ (28) which becomes the Fuchs equation with three canonical singularities
$ a_1 $ ,$ a_2 $ and$ a_3 $ , as follows:$ \begin{aligned}[b] \phi''(z) &+ \left( \frac{1-\bar{\alpha}_1-\bar{\alpha}_2}{z-a_1} + \frac{1-\bar{\beta}_1-\bar{\beta}_2}{z-a_2} + \frac{1-\bar{\gamma}_1-\bar{\gamma}_2}{z-a_3} \right) \phi'(z) + \left( \frac{\bar{\alpha}_1 \bar{\alpha}_2 (a_1-a_2) (a_1-a_3)}{z-a_1} + \frac{\bar{\beta}_1 \bar{\beta}_2 (a_2-a_3) (a_2-a_1)}{z-a_2}\right. \\&\left.+ \frac{\bar{\gamma}_1 \bar{\gamma}_2 (a_3-a_1) (a_3-a_2)}{z-a_3} \right) \frac{\phi(z)}{(z-a_1)(z-a_2)(z-a_3)} = 0, \end{aligned} $
(29) where
$ a_1 = 0 $ ,$ a_2 = -1 $ , and$ a_3 = 1 $ and$ \begin{aligned}[b]& \bar{\alpha}_1 = \frac12 \pm \nu, \quad \bar{\alpha}_2 = \frac12 \mp \nu, \quad \bar{\beta}_1 = - \bar{\beta}_2 = \pm {\rm i} \frac{w \xi_{\mathrm o} - 2 q_\mathrm{eff} \ell}2, \\& \bar{\gamma}_1 = - \bar{\gamma}_2 = \pm {\rm i} \frac{w \xi_{\mathrm o}}2, \end{aligned}$
(30) and
$ \bar{\alpha}_1 + \bar{\alpha}_2 + \bar{\beta}_1 + \bar{\beta}_2 + \bar{\gamma}_1 + \bar{\gamma}_2 = 1 $ is satisfied. The Fuchs Eq. (29) can be transformed into the standard hypergeometric function$ \zeta (1 - \zeta) \psi''(\zeta) + \left[ \tilde{\gamma} - (1 + \tilde{\alpha} + \tilde{\beta}) \zeta \right] \psi'(\zeta) - \tilde{\alpha} \tilde{\beta} \psi(\zeta) = 0, $
(31) via the conformal coordinate transformation
$ \zeta = \frac{(a_2-a_3)(z-a_1)}{(a_2-a_1)(z-a_3)}, \quad {\rm{}} \;\; \phi(z) = \left(\frac{z-a_1}{z-a_3}\right)^{\bar{\alpha}_1} \left(\frac{z-a_2}{z-a_3}\right)^{\bar{\beta}_1} \psi(\zeta), $
(32) where
$ \tilde{\alpha} = \bar{\alpha}_1 + \bar{\beta}_1 + \bar{\gamma}_1, \tilde{\beta} = \bar{\alpha}_1 + \bar{\beta}_1 + \bar{\gamma}_2 $ and$ \tilde{\gamma} = 1 + \bar{\alpha}_1 - \bar{\alpha}_2 $ . (Note that one can freely choose the indices$ i = 1,\; 2 $ for$ \bar{\alpha}_i $ ,$ \bar{\beta}_i $ and$ \bar{\gamma}_i $ .)For Eq. (28), we have
$ \zeta = 2z/(z-1) $ ,$\tilde{\alpha} = \dfrac12 \pm \nu + $ $ {\rm i} w \xi_{\mathrm o} - {\rm i} q_\mathrm{eff} \ell, \;\;\;\;\tilde{\beta} = \dfrac12 \pm \nu - {\rm i} q_\mathrm{\rm eff} \ell, \;\;\;\; \tilde{\gamma} = 1 \pm 2\nu$ . Therefore, the explicit solutions in the near horizon near extreme region are$ \begin{aligned}[b] \phi(z) =& c_1 \left(\frac{z}{z-1}\right)^{\frac12 + \nu} \left(\frac{z+1}{z-1}\right)^{{\rm i} \frac{w \xi_{\mathrm o}}{2} - {\rm i} q_\mathrm{eff} \ell} {_2F_1}\left(\frac12 + \nu + {\rm i} w \xi_{\mathrm o} - {\rm i} q_\mathrm{eff} \ell, \frac12 + \nu - {\rm i} q_\mathrm{eff} \ell; 1 + 2 \nu; \frac{2z}{z-1} \right) \\ & + c_2 \left(\frac{z}{z-1}\right)^{\frac12 - \nu} \left(\frac{z+1}{z-1}\right)^{{\rm i} \frac{w \xi_{\mathrm o}}{2} - {\rm i} q_\mathrm{eff} \ell} {_2F_1}\left(\frac12 - \nu + {\rm i} w \xi_{\mathrm o} - {\rm i} q_\mathrm{eff} \ell, \frac12 - \nu - {\rm i} q_\mathrm{eff} \ell; 1 - 2 \nu; \frac{2z}{z-1} \right).\\ \end{aligned} $ (33) -
At the horizon of the AdS2 black brane,
$ z = 1 $ , Eq. (33) is expanded as follows:$ \phi(z) = c_H^{(\mathrm {in})}(1-z)^{-{\rm i} \frac{w}{4\pi T_{ n}}} + c_H^{(\mathrm {out})} (1-z)^{{\rm i} \frac{w}{4\pi T_{n}}}, $
(34) where
$ \begin{aligned}[b] c_H^{(\mathrm {in})} = c_1 (-)^{-\frac12 - \nu - {\rm i} \frac{w}{2 \pi T_{ n}} + {\rm i} q_\mathrm{eff} \ell} \, 2^{-\frac12 - \nu + {\rm i} \frac{w}{4\pi T_{ n}}} \frac{\Gamma\left(1 + 2\nu\right) \Gamma\left({\rm i} \dfrac{w}{2\pi T_{ n}}\right)}{\Gamma\left(\dfrac12 + \nu + {\rm i} q_\mathrm{eff} \ell\right) \Gamma\left(\dfrac12 + \nu + {\rm i} \dfrac{w}{2\pi T_{ n}} - {\rm i} q_\mathrm{eff} \ell\right)}\end{aligned} $ $ \begin{aligned}[b]+ c_2 (-)^{-\frac12 + \nu - {\rm i} \frac{w}{2 \pi T_{ n}} + {\rm i} q_\mathrm{eff} \ell} \, 2^{-\frac12 + \nu + {\rm i} \frac{w}{4\pi T_{ n}}} \frac{\Gamma\left(1 - 2\nu\right) \Gamma\left({\rm i} \dfrac{w}{2\pi T_{ n}}\right)}{\Gamma\left(\dfrac12 - \nu + {\rm i} q_\mathrm{eff} \ell\right) \Gamma\left(\dfrac12 - \nu + {\rm i} \dfrac{w}{2\pi T_{ n}} - {\rm i} q_\mathrm{eff} \ell\right)}, \end{aligned} $
(35) and
$ \begin{aligned}[b] c_H^{(\mathrm {out})} =& c_1 (-)^{-\frac12 - \nu - {\rm i} \frac{w}{2 \pi T_{ n}} + {\rm i} q_\mathrm{eff} \ell} \, 2^{-\frac12 - \nu - {\rm i} \frac{w}{4 \pi T_{ n}}} \frac{\Gamma\left(1 + 2\nu\right) \Gamma\left(-{\rm i} \dfrac{w}{2\pi T_{ n}}\right)}{\Gamma\left(\dfrac12 + \nu - {\rm i} q_\mathrm{eff} \ell\right) \Gamma\left(\dfrac12 + \nu - {\rm i} \dfrac{w}{2\pi T_{n}} + {\rm i} q_\mathrm{eff} \ell\right)} \\ &+ c_2 (-)^{-\frac12 + \nu - {\rm i} \frac{w}{2 \pi T_{ n}} + {\rm i} q_\mathrm{eff} \ell} \, 2^{-\frac12 + \nu - {\rm i} \frac{w}{4\pi T_{ n}}} \frac{\Gamma\left(1 - 2\nu\right) \Gamma\left(-{\rm i} \dfrac{w}{2\pi T_{ n}}\right)}{\Gamma\left(\dfrac12 - \nu - {\rm i} q_\mathrm{eff} \ell\right) \Gamma\left(\dfrac12 - \nu - {\rm i} \dfrac{w}{2\pi T_{ n}} + {\rm i} q_\mathrm{eff} \ell\right)}. \end{aligned} $
(36) In contrast, at the AdS2 boundary,
$ z \rightarrow 0 $ , the asymptotic expansion of Eq. (33) is$ \begin{aligned}[b] \phi(z) =& c_2 (-)^{\frac12 - \nu + {\rm i} \frac{w}{4\pi T_{ n}} - {\rm i} q_\mathrm{eff} \ell} z^{\frac{1}{2}- \nu} + c_1 (-)^{\frac12 + \nu + i \frac{w}{4\pi T_{\mathrm n}} - {\rm i} q_\mathrm{eff} \ell} z^{\frac{1}{2}+ \nu} \\=& {\cal A}(w, \vec{k})z^{\frac{1}{2}- \nu}+{\cal B}(w, \vec{k})z^{\frac{1}{2}+ \nu},\end{aligned} $
(37) where
$ {\cal A} $ is the source of the charged scalar field in the bulk AdS2, while$ {\cal B} $ is the response or the operator$ {\cal \hat{O}}(w,\vec{k}) $ (in the momentum space) of the boundary CFT1 (i.e., the IR CFT) dual to the charged scalar field in the bulk AdS2 background. Note that in order to obtain the propagating modes,$ \nu $ should be purely imaginary, which can be set as$ \nu \equiv {\rm i} |\nu| $ , i.e.,$ \phi(z) = c_B^{(\mathrm {out})} z^{\frac{1}{2} - {\rm i}|\nu|} + c_B^{(\mathrm {in})} z^{\frac{1}{2} + {\rm i}|\nu|} $ . It was shown in [3] that the condition of an imaginary$ \nu $ is equivalent to the violation of the BF bound in AdS2 spacetime, namely$ \begin{aligned} \tilde{m}_\mathrm{eff}^2 < -\frac{1}{4\ell^2}, \end{aligned} $
(38) which corresponds to a complex conformal weight of the scalar operator in the dual IR CFT.
-
The Schwinger pair production rate
$ |\mathfrak{b}|^2 $ and the absorption cross section ratio$ \sigma_{\mathrm{abs}} $ can be calculated from the radial flux by imposing different boundary conditions$ \begin{align} {\cal F} = {\rm i} \left(\frac{r_{\mathrm o}}{L}\right)^{d-1} (1-z^2) (\Phi \partial_z \Phi^* - \Phi^* \partial_z \Phi), \end{align} $
(39) which gives
$ \begin{aligned}[b] {\cal F}_B^{(\mathrm {in})} =& 2 |\nu| \left(\frac{r_{\mathrm o}}{L}\right)^{d-1} |c_B^{(\mathrm {in})}|^2,\\ {\cal F}_B^{(\mathrm {out})} =& -2 |\nu| \left(\frac{r_{\mathrm o}}{L}\right)^{d-1} |c_B^{(\mathrm {out})}|^2, \\ {\cal F}_H^{(\mathrm {in})} =& \frac{w}{2\pi T_{\mathrm n}} \left(\frac{r_{\mathrm o}}{L}\right)^{d-1} |c_H^{(\mathrm {in})}|^2, \\ {\cal F}_H^{(\mathrm {out})} =& -\frac{w}{2\pi T_{\mathrm n}} \left(\frac{r_{\mathrm o}}{L}\right)^{d-1} |c_H^{(\mathrm {out})}|^2, \end{aligned} $
(40) where
$ {\cal F}_B^{(\mathrm {in})} $ and$ {\cal F}_B^{(\mathrm {out})} $ are the ingoing and outgoing fluxes at the AdS2 boundary, while$ {\cal F}_H^{(\mathrm {in})} $ and$ {\cal F}_H^{(\mathrm {out})} $ are the ingoing and outgoing fluxes at the AdS2 black brane horizon, respectively.The Schwinger pair production rate
$ {{{\left| {{\mathfrak{b}^{{\text{Ad}}{{\text{S}}_{\text{2}}}}}} \right|}^{\text{2}}}} $ can be computed either by choosing the inner boundary condition or the outer boundary condition, which gives the same result [3], e.g., by adopting the outer boundary condition, i.e.,$ {\cal F}_B^{(\mathrm {in})} = 0 $ , ($ c_B^{(\mathrm {in})} = 0 \Rightarrow c_1 = 0 $ ),$ \begin{aligned}[b] {{{\left| {{\mathfrak{b}^{{\text{Ad}}{{\text{S}}_{\text{2}}}}}} \right|}^{\text{2}}}}& = \frac{{\cal F}_B^{(\mathrm {out})}}{{\cal F}_H^{(\mathrm {in})}} = \frac{4 \pi T_{ n} |\nu|}{w} \left|\frac{c_B^{(\mathrm {out})}}{c_H^{(\mathrm {in})}}\right|^2 = \frac{ 8 \pi T_{ n} |\nu|}{w} \left| \frac{\Gamma\left(\dfrac12 - {\rm i}|\nu| + {\rm i} q_\mathrm{eff} \ell\right) \Gamma\left(\dfrac12 - {\rm i}|\nu| + {\rm i} \dfrac{w}{2\pi T_{ n}} -{\rm i} q_\mathrm{eff} \ell\right)}{\Gamma\left(1 - 2{\rm i}|\nu|\right) \Gamma\left({\rm i} \dfrac{w}{2\pi T_{ n}}\right)}\right|^2\\ & = \frac{2\sinh\left(2\pi|\nu| \right)\sinh\left(\dfrac{w}{2T_{ n}}\right)}{\cosh\pi\left(|\nu|-q_\mathrm{eff} \ell\right)\cosh\pi\left(|\nu|-\dfrac{w}{2\pi T_{ n}}+q_\mathrm{eff} \ell\right)}. \end{aligned} $ (41) Similarly, by adopting the outer boundary condition, the absorption cross section ratio is computed as
$ \begin{aligned}[b] \sigma _{{\text{abs}}}^{{\text{Ad}}{{\text{S}}_{\text{2}}}}& = \frac{{\cal F}_B^{(\mathrm {out})}}{{\cal F}_H^{(\mathrm {out})}} = \frac{4 \pi T_{ n} |\nu|}{w} \left|\frac{c_B^{(\mathrm {out})}}{c_H^{(\mathrm {out})}}\right|^2 = \frac{ 8 \pi T_{ n} |\nu|}{w} \left| \frac{\Gamma\left(\dfrac12 - {\rm i}|\nu| - {\rm i} q_\mathrm{eff} \ell\right) \Gamma\left(\dfrac12 - {\rm i}|\nu| - {\rm i} \dfrac{w}{2\pi T_{ n}} +{\rm i} q_\mathrm{eff} \ell\right)}{\Gamma\left(1 - 2{\rm i}|\nu|\right) \Gamma\left(-{\rm i} \dfrac{w}{2\pi T_{ n}}\right)}\right|^2\\ & = \frac{2\sinh\left(2\pi|\nu| \right)\sinh\left(\dfrac{w}{2T_{ n}}\right)}{\cosh\pi\left(|\nu|+q_\mathrm{eff} \ell\right)\cosh\pi\left(|\nu|+\dfrac{w}{2\pi T_{ n}}-q_\mathrm{eff} \ell\right)}. \end{aligned} $
(42) The pair production rate and the absorption cross section ratio are connected by the simple relation
$ {{{\left| {{\mathfrak{b}^{{\text{Ad}}{{\text{S}}_{\text{2}}}}}} \right|}^{\text{2}}}} =-\sigma_{\mathrm{abs}}(|\nu|\rightarrow -|\nu|). $
(43) It was shown that the abovementioned relation also holds for a charged scalar field [11] and for a charged spinor field [4], both in a four-dimensional near extremal RN black hole.
-
The two-point retarded Green's function of the boundary operator dual to the bulk charged scalar field is computed through
$ \begin{aligned}[b] G_R^{\mathrm {AdS_2}}(w, \vec{k}) \equiv & \langle {\cal \hat{O}} {\cal \hat{O}} \rangle_R\\ =& -2 {\cal F}|_{z\rightarrow 0} \sim \frac{{\cal B}(w, \vec{k})}{{\cal A}(w, \vec{k})} \\&+ \rm{contact\; terms} \end{aligned}$
(44) by taking the inner boundary condition, i.e.,
$ {\cal F}_H^{(\mathrm {out})} = 0 $ , which gives$ \frac{c_2}{c_1} = (-)^{1 - 2\nu} \, 2^{-2\nu} \frac{\Gamma\left(1 + 2\nu\right) \Gamma\left(\dfrac12 - \nu - {\rm i} q_\mathrm{eff} \ell\right) \Gamma\left(\dfrac12 - \nu - {\rm i} \dfrac{w}{2\pi T_{\mathrm n}} + {\rm i} q_\mathrm{eff} \ell\right)}{\Gamma\left(1 - 2\nu\right) \Gamma\left(\dfrac12 + \nu - {\rm i} q_\mathrm{eff} \ell\right) \Gamma\left(\dfrac12 + \nu - {\rm i} \dfrac{w}{2\pi T_{\mathrm n}} + {\rm i} q_\mathrm{eff} \ell\right)}. $ (45) Thus, the two-point retarded Green's function is
$ G_R^{\mathrm {AdS_2}}(w, \vec{k}) \sim \frac{{\cal B}(\omega, \vec{k})}{{\cal A}(\omega, \vec{k})} = (-)^{2\nu} \frac{c_1}{c_2} = (-)^{4\nu-1} \, 2^{2\nu} \frac{\Gamma\left(1 - 2\nu\right) \Gamma\left(\dfrac12 + \nu - {\rm i} q_\mathrm{eff} \ell\right) \Gamma\left(\dfrac12 + \nu - {\rm i} \dfrac{w}{2\pi T_{\mathrm n}} + {\rm i} q_\mathrm{eff} \ell\right)}{\Gamma\left(1 + 2\nu\right) \Gamma\left(\dfrac12 - \nu - {\rm i} q_\mathrm{eff} \ell\right) \Gamma\left(\dfrac12 - \nu - {\rm i} \dfrac{w}{2\pi T_{\mathrm n}} + {\rm i} q_\mathrm{eff} \ell\right)}. $
(46) In addition, the corresponding boundary condition (
$ {\cal F}_B^{(\mathrm {in})} = 0 $ and$ {\cal F}_H^{(\mathrm {out})} = 0 $ ) is used to obtain the quasinormal modes of the charged scalar field in AdS2 spacetime, which correspond to the poles of the retarded Green's function of dual operators (with complex conformal weight$ h_R = \dfrac12+\nu $ ) in the IR CFT, namely$ \begin{aligned}[b]& \frac12 +\nu - {\rm i} \frac{w}{2\pi T_{ n}} + {\rm i} q_\mathrm{eff} \ell = -N \Rightarrow w\\ =& 2\pi T_{ n}\left(q_\mathrm{eff} \ell-{\rm i}N-{\rm i}h_R\right),\quad N = 0,1,\cdots. \end{aligned} $
(47) Eq. (47) gives the quasinormal modes of the charged scalar field perturbation.
-
In this section, we describe our study of the pair production for the whole spacetime of RN-AdS5. Like before, we need to solve the corresponding radial Klein equation for the scalar field.
To find the solution in the full region, we focus on
$ d = 4 $ and the near extremal cases. By introducing the coordinate transformation$ \varrho = \dfrac{{{r^2}}}{{M'}} $ (and denoting$ M'{\text{ = }} $ $ {{\text{G}}_{d + 1}}{L^2}M $ ,$ \varrho_{\mathrm o} = \dfrac{r_{\mathrm o}^2}{M'} $ and$ \varrho _* = \dfrac{r_*^2}{M'} $ ), the radial Eq. (18) can be expressed as$ \phi ''(\varrho ) + \left( {\frac{1}{{\varrho - {\varrho _1}}} + \frac{1}{{\varrho - {\varrho _2}}} + \frac{1}{{\varrho - {\varrho_{\mathrm o} }}}} \right)\phi '\left( \varrho \right) + \left( {\frac{{\varrho {{\left( {\tilde \omega \varrho - \tilde q\mu {\varrho_{\mathrm o} }} \right)}^2}}}{{{{\left( {\varrho - {\varrho _1}} \right)}^2}{{\left( {\varrho - {\varrho _2}} \right)}^2}{{\left( {\varrho - {\varrho_{\mathrm o} }} \right)}^2}}} - \frac{{{{\tilde m}^{\rm{2}}}\varrho + {{\tilde k}^2}}}{{\left( {\varrho - {\varrho _1}} \right)\left( {\varrho - {\varrho _2}} \right)\left( {\varrho - {\varrho_{\mathrm o} }} \right)}}} \right)\phi \left( \varrho \right) = 0, $ (48) where the parameters are
$ \tilde \omega = \dfrac{{{L^2}(\omega + q\mu )}}{{2\sqrt {M'} }} $ ,$ \tilde q = \dfrac{{{L^2}q}}{{2\sqrt {M'} }} $ ,$ \tilde m = \dfrac{Lm}{2} $ ,$ \tilde k = \dfrac{{{L^2}\left| {\vec k} \right|}}{{2\sqrt {M'} }} $ and$ \begin{aligned}[b] \varrho _1 & = - \frac{1}{2}{\varrho_{\mathrm o} } - \frac{1}{2}\sqrt {{\varrho_{\mathrm o} }^{\text{2}} + 8\frac{{{\varrho _*}^{\text{3}}}}{{{\varrho_{\mathrm o} }}}}, \\ \varrho _2 & = - \frac{1}{2}{\varrho_{\mathrm o} }{\text{ + }}\frac{1}{2}\sqrt {{\varrho_{\mathrm o} }^{\text{2}} + 8\frac{{{\varrho _*}^{\text{3}}}}{{{\varrho_{\mathrm o} }}}}. \end{aligned} $
(49) Further, defining another coordinate
$ y\equiv \frac{{\varrho - {\varrho_{\mathrm o} }}}{{{\varrho_{\mathrm o} }}},\quad a\equiv \frac{\varrho _2-\varrho _{\mathrm{o}}}{\varrho _{\mathrm{o}}},\quad {} b\equiv \frac{\varrho _1-\varrho _{\mathrm{ o }}}{\varrho _{\mathrm{ o }}}, $
(50) the metric of the RN-AdS5 black hole becomes
$ \begin{aligned}[b] {\rm d}s^2& = \frac{L^2 {\rm d}y^2}{4(1+y)^2 f(y)}+\frac{r_{\mathrm{o}}^2}{L^2}(1+y)\left(-f(y){\rm d}t^2+{\rm d}x_i^2 \right),\\ A& = \frac{\mu y}{1+y}{\rm d}t, \end{aligned} $
(51) where
$\begin{aligned}[b] f(y) =& 1-\frac{M'}{r_{\mathrm{o}}^4}(1+y)^{-2}+\frac{Q'^2}{r_{\mathrm{o}}^6}(1+y)^{-3}, \\ {{Q'}^2} =& {{\text{G}}_{d + 1}}{L^2}{Q^2}. \end{aligned}$
(52) Moreover, Eq. (48) transforms into
$\begin{aligned}[b] \phi ''(y) &+ \left( {\frac{1}{y} + \frac{1}{{y - a}} + \frac{1}{{y - b}}} \right)\phi '\left( y \right) \\&+ \left( \frac{{{{\left( {\tilde \omega (y + 1) - \tilde q\mu } \right)}^2}(y + 1)}}{{{y^2}{{\left( {y - a} \right)}^2}{{\left( {y - b} \right)}^2}}}\right.\\&\left. - \frac{{{{\tilde m}^2}(y + 1){\varrho_{\mathrm o} } + {{\tilde k}^2}}}{{y\left( {y - a} \right)\left( {y - b} \right)}} \right)\frac{{\phi \left( y \right)}}{{{\varrho_{\mathrm o} }}} = 0.\end{aligned}$
(53) To solve
$ \phi(y) $ , first, we determine its exponents at the corresponding singularities$ 0 $ , a, b, and$ \infty $ , which are$ \alpha _{1,2} $ ,$ \beta _{1,2} $ ,$ \gamma _{1,2} $ , and$ \delta _{1,2} $ , respectively,$ \begin{aligned}[b] &\alpha _{1,2} = \pm {\rm i}\frac{(\tilde \omega - \tilde q\mu) }{ab\sqrt {\varrho_{\mathrm o}}} = \pm\frac{{\rm i}\omega}{4\pi T},\\ &\beta _{1,2} = \pm {\rm i}\frac{{(\tilde \omega (1 + b) - \tilde q\mu )\sqrt {1 + b} }}{{(a - b)b\sqrt {{\varrho_{\mathrm o} }} }},\\& \gamma _{1,2} = \pm {\rm i}\frac{{(\tilde \omega (1 + a) - \tilde q\mu )\sqrt {1 + a} }}{{(b - a)a\sqrt {{\varrho_{\mathrm o} }} }}, \end{aligned} $
(54) where the index “1” corresponds to the “
$ + $ ” sign, and the index “2” corresponds to the “$ - $ ” sign. Then, decomposing$ \phi \left( y \right) $ as$ \phi \left( y \right) = {\left( {\frac{y}{{y - b}}} \right)^{{\alpha _1}}}{\left( {\frac{{y - a}}{{y - b}}} \right)^{{\gamma _1}}}R(y), $
(55) we obtain
$ \begin{aligned}[b] R''(y) &+ \left( {\frac{1}{y} + \frac{1}{{y - a}} + \frac{1}{{y - b}} - \frac{{2b{\alpha _1}}}{{y(y - b)}} + \frac{{2(a - b){\gamma _1}}}{{(y - a)(y - b)}}} \right)R'\left( y \right) \\&+ {V_2}R(y) = 0, \end{aligned} $
(56) where
$ {V_2} \equiv - \frac{{(2 + 3a - {a^2}y){{\tilde \omega }^{\rm{2}}} - {\rm{4}}\left( {a + 1} \right)\tilde \omega \tilde q\mu {\rm{ + }}\left( {a + {\rm{2}}} \right){{\tilde q}^{\rm{2}}}{\mu ^{\rm{2}}}}}{{{\varrho _{\rm{o}}}{a^2}y\left( {y - a} \right){{\left( {y - b} \right)}^2}}} - {M_1}, $
(57) and
$\begin{aligned}[b] M_1 =& \frac{{\left( {b - a} \right){\gamma _1}}}{{y\left( {y - a} \right)(y - b)}} + \frac{{2b\left( {a - b} \right){\alpha _1}{\gamma _1}}}{{y\left( {y - a} \right){{(y - b)}^2}}} + \frac{{{{\tilde m}^2}(y + 1){\varrho_{\mathrm o} } + {{\tilde k}^2}}}{{{\varrho_{\mathrm o} }y\left( {y - a} \right)\left( {y - b} \right)}}\\& + \frac{{b{\alpha _1}}}{{y\left( {y - a} \right)(y - b)}}. \end{aligned}$
(58) -
We divide the regions into a near region
$ y = \frac{{\varrho - {\varrho_{\mathrm o} }}}{{{\varrho_{\mathrm o} }}} \ll 1, $
(59) and a far region
$ y = \frac{{\varrho - {\varrho_{\mathrm o} }}}{{{\varrho_{\mathrm o} }}}\gg -a, $
(60) and an overlapping region, in which
$ -a \ll 1. $
(61) The physical reasoning of
$ -a \ll 1 $ relies on the observation that the temperature of a black hole is$ T = \frac{{{r_{\mathrm o} }}}{{\pi {L^2}}}\left( {1 - \frac{{{\varrho _*}^3}}{{{\varrho_{\mathrm o} }^3}}} \right), $
(62) which gives
$ -a \to 0 $ for$ T \to 0 $ , as$ - a = \frac{3}{2} - \frac{1}{2}\sqrt {9 - \frac{8\pi L^2 T}{r_{\mathrm o} }}. $
(63) We want to point out that the matching condition in Eq. (61) indicates that the near extremal condition is essential for matching the solutions in the near and far regions. It is not necessary for the frequency to be infinitely small; however, the frequency should definitely not be very large compared with the temperature T; otherwise, the backreaction to the background geometry cannot be ignored.
Now we find the approximate solutions in different regions. First, by using the near region condition (
$ y \ll 1 $ ), Eq. (53) reduces to$\begin{aligned}[b] \phi ''(y) &+ \left( {\frac{1}{y} + \frac{1}{{y - a}}} \right)\phi '\left( y \right) \\&+ \left( { - \frac{{{{(a{\alpha _1} - {\rm i}{q_{{\rm{eff}}}}\ell y)}^2}}}{{{y^{\rm{2}}}{{\left( {y - a} \right)}^{\rm{2}}}}} + \frac{{{{\tilde m}^2}{\varrho _{\rm{o}}} + {{\tilde k}^2}}}{{{\varrho _{\rm{o}}}by(y - a)}}} \right)\phi (y) = 0. \end{aligned} $
(64) Obviously Eq. (64) can be solved by the hypergeometric function as
$\begin{aligned}[b] \phi \left( y \right) =& {\left( {y - a} \right)^{{\alpha _1} - {\rm i}{q_{{\rm{eff}}}}\ell}}\bigg( {{c_3}{y^{{\alpha _1}}}_2{F_1}\left( {\alpha ,\beta ;\gamma ;\frac{y}{a}} \right)} \\&+ {c_4}{y^ - }^{{\alpha _1}}{}_2{F_1}\left( {1 - \gamma + \alpha ,1 - \gamma + \beta ;2 - \gamma ;\frac{y}{a}} \right) \bigg), \end{aligned}$
(65) where
$\alpha = \dfrac{1}{2} + \nu + 2{\alpha _1} - {\rm i}{q_{{\rm{eff}}}}\ell$ ,$\beta = \dfrac{1}{2} - \nu + 2{\alpha _1} - {\rm i}{q_{{\rm{eff}}}}\ell$ , and$ \gamma = 1 + 2{\alpha _1} $ , and$ \ell = L/\sqrt{12} $ is the radius of the effective AdS2 geometry in the near horizon region of the RN-AdS5 black hole. Second, in the far region, by using the condition ($ y \gg -a $ ), Eq. (56) can turn into$ R''(y) + \left( {\frac{2}{y} + \frac{1}{{y - b}} - \frac{{4b{\alpha _1}}}{{y(y - b)}}{\rm{ + }}\frac{{2{\rm i}b{q_{{\rm{eff}}}}\ell }}{{y(y - b)}}} \right)R'\left( y \right) + {V_3}R(y) = 0, $
(66) where
$ \begin{aligned}[b] {V_3} \equiv & \frac{{{{\tilde \omega }^2}}}{{{\varrho _{\rm{o}}}y{{(y - b)}^2}}} + \frac{{4{b^2}{\alpha _1}\left( {{\alpha _1} - {\rm i}{q_{{\rm{eff}}}}\ell } \right)}}{{{y^2}{{(y - b)}^2}}} \\&- \frac{{b\left( {2{\alpha _1} - {\rm i}{q_{{\rm{eff}}}}\ell } \right)}}{{{y^2}(y - b)}} - \frac{{{{\tilde m}^2}(y + 1){\varrho _{\rm{o}}} + {{\tilde k}^2}}}{{{\varrho _{\rm{o}}}{y^2}(y - b)}},\end{aligned} $
(67) (where the relation
$ {\gamma _1} \approx {\alpha _1} + \dfrac{{{\rm i}\tilde q\mu }}{{b\sqrt {{\varrho _{\rm{o}}}} }} = {\alpha _1} - {\rm i}{q_{{\rm{eff}}}}\ell $ when$ \left| a \right| \ll 1 $ is used). Similarly, Eq. (66) has a solution in terms of the hypergeometric function as$\begin{aligned}[b] \phi \left( y \right) =& {\left( {\frac{y}{b} - 1} \right)^\lambda}\bigg( {{c_5}{y^{\nu - \frac{1}{2}}}_2{F_1}\left( {\alpha ',\beta ';\gamma ';\frac{y}{b}} \right)} \\&+ {c_6}{y^{ - \nu - \frac{1}{2}}}_2{F_1}\left( {1 - \gamma ' + \alpha ',1 - \gamma ' + \beta ';2 - \gamma ';\frac{y}{b}} \right) \bigg) \end{aligned}$
(68) in which
$ \alpha ' = \dfrac{1}{2} + \nu + \Delta + \lambda $ ,$ \beta ' = \dfrac{1}{2} + \nu - \Delta + \lambda $ ,$ \gamma ' = 1 + 2\nu $ ,$ \Delta = \sqrt {1 + {{\tilde m}^2}} $ , and$\lambda = \sqrt {{{\left( {{\rm i}{q_{{\rm{eff}}}}\ell } \right)}^2} - \dfrac{{{{\tilde \omega }^2}}}{{{\varrho _{\rm{o}}}b}}}$ . -
In the overlapping region, one has the inequalities
$- a \ll y \ll 1 < -b$
(69) (
$ 1 < - b $ since$ - b = \dfrac{3}{2}{\text{ + }}\dfrac{1}{2}\sqrt {9 - \dfrac{{8\pi {L^2}T}}{{{r_{\mathrm o} }}}} \to {\text{3}} $ , as$ T \to {\text{0}} $ ), which means$ \left| {\dfrac{a}{y}} \right| \to 0 $ and$ \left| {\dfrac{y}{b}} \right| \to 0 $ , which transforms Eqs. (65) and (68) into the following forms: the near regionsolution$ \begin{aligned}[b] \phi (y) =& \left( {{\left( { - 1} \right)}^{ - \alpha }}{c_3}\frac{{\Gamma \left( \gamma \right)\Gamma \left( {\beta - \alpha } \right)}}{{\Gamma \left( \beta \right)\Gamma \left( {\gamma - \alpha } \right)}}\right.\\&\left. + {{\left( { - 1} \right)}^{ - 1 + \gamma - \alpha }}{c_4}\frac{{\Gamma \left( {2 - \gamma } \right)\Gamma \left( {\beta - \alpha } \right)}}{{\Gamma \left( {1 - \gamma + \beta } \right)\Gamma \left( {1 - \alpha } \right)}} \right){y^{ - \frac{1}{2} - \nu }} \\ &+ \left( {{\left( { - 1} \right)}^{ - \beta }}{c_3}\frac{{\Gamma \left( \gamma \right)\Gamma \left( {\alpha - \beta } \right)}}{{\Gamma \left( \alpha \right)\Gamma \left( {\gamma - \beta } \right)}} \right.\\&+\left. {{\left( { - 1} \right)}^{ - 1 + \gamma - \beta }}{c_4}\frac{{\Gamma \left( {2 - \gamma } \right)\Gamma \left( {\alpha - \beta } \right)}}{{\Gamma \left( {1 - \gamma + \alpha } \right)\Gamma \left( {1 - \beta } \right)}} \right){y^{ - \frac{1}{2} + \nu }} \end{aligned} $
(70) and the far region solution
$ \phi (y) \to {\left( { - 1} \right)^\lambda}{c_5}{y^{ - \frac{1}{2} + \nu }} + {\left( { - 1} \right)^\lambda}{c_6}{y^{ - \frac{1}{2} - \nu }}. $
(71) Comparing these two identities, one finds the connection relations
$ \begin{aligned}[b] {c_5} =& {( - 1)^{ - \lambda}}\left( {{\left( { - 1} \right)}^{ - \beta }}{c_3}\frac{{\Gamma \left( \gamma \right)\Gamma \left( {\alpha - \beta } \right)}}{{\Gamma \left( \alpha \right)\Gamma \left( {\gamma - \beta } \right)}}\right.\\&\left. + {{\left( { - 1} \right)}^{ - 1 + \gamma - \beta }}{c_4}\frac{{\Gamma \left( {2 - \gamma } \right)\Gamma \left( {\alpha - \beta } \right)}}{{\Gamma \left( {1 - \gamma + \alpha } \right)\Gamma \left( {1 - \beta } \right)}} \right), \end{aligned}$
(72) $ \begin{aligned}[b]{c_6} =& {( - 1)^{ - \lambda}}\left( {{\left( { - 1} \right)}^{ - \alpha }}{c_3}\frac{{\Gamma \left( \gamma \right)\Gamma \left( {\beta - \alpha } \right)}}{{\Gamma \left( \beta \right)\Gamma \left( {\gamma - \alpha } \right)}}\right.\\&\left. + {{\left( { - 1} \right)}^{ - 1 + \gamma - \alpha }}{c_4}\frac{{\Gamma \left( {2 - \gamma } \right)\Gamma \left( {\beta - \alpha } \right)}}{{\Gamma \left( {1 - \gamma + \beta } \right)\Gamma \left( {1 - \alpha } \right)}} \right).\end{aligned} $
(73) -
Now we denote the radial flux of the charged scalar field in metric (51) as
$ {\cal D} $ :$ {\cal D} = \frac{2{\rm i}r_{\mathrm{o}}^4(1+y)^3 f(y)}{L^5}\bigg(\phi(y)\partial_y \phi^*(y)- \phi^*(y)\partial_y \phi(y) \bigg). $
(74) In the near horizon limit, i.e.,
$ y \to 0 $ , Eq. (65) reduces to$ \phi (y) = {c_3}{y^{{\alpha _1}}}{\left( {y - a} \right)^{{\alpha _1} - {\rm i}{q_{{\rm{eff}}}}\ell }} + {c_4}{y^{ - {\alpha _1}}}{\left( {y - a} \right)^{{\alpha _1} - {\rm i}{q_{{\rm{eff}}}}\ell }}, $
(75) where the first part is the outgoing mode, and the second part is the ingoing mode. Further, the asymptotic form of
$ \phi(y) $ at the boundary ($ y\to \infty $ ) of the AdS5 spacetime results in the form$ \phi(y) = A(\tilde{\omega}, \tilde{k})y^{- 1 + \Delta}+B(\tilde{\omega}, \tilde{k})y^{- 1 - \Delta}, $
(76) where
$ A(\tilde{\omega}, \tilde{k}) $ is the source of the charged scalar field in the bulk RN-AdS5 black hole, while$ B(\tilde{\omega}, \tilde{k}) $ is the response (the operator) of the boundary CFT4 (i.e., the UV CFT) dual to the charged scalar field in the bulk. As in the case of the AdS2 spacetime, the condition for the propagating modes requires an imaginary$ \Delta $ , i.e.,$ \Delta = {\rm i}\left| \Delta \right| $ , which means$ m^2\leqslant -\frac{4}{L^2}, $
(77) namely, the violation of the BF bound in AdS5 spacetime.
Therefore, the corresponding outgoing and ingoing fluxes at the horizon and the boundary of the near extremal RN-AdS5 black brane are
$ \begin{aligned}[b] {\cal D}_H^{(\mathrm{out})} & = \frac{{4\pi \omega T{r_{\mathrm{o}}}^2}}{{abL}}|{c_3}{|^2} = \frac{2r_{\mathrm{o}}^3\omega}{L^3}|c_3|^2 ,\\{\kern 1pt} {\cal D}_H^{(\mathrm{in})} &= - \frac{{4\pi \omega T{r_{\rm{o}}}^2}}{{abL}}|{c_4}{|^2} = -\frac{2r_{\mathrm{o}}^3\omega}{L^3}|c_4|^2, \\ {\cal D}_B^{({\mathrm{out}})} & = \frac{{4|\Delta |{r_{\rm{o}}}^4}}{{{L^5}}}\left| A( {\tilde \omega ,\tilde k})\right|^2,\\ {\cal D}_B^{({\mathrm{in}})}& = - \frac{{4|\Delta |{r_{\rm{o}}}^4}}{{{L^5}}}\left|B( {\tilde \omega ,\tilde k})\right|^2. \end{aligned} $
(78) The absorption cross section ratio
$ \sigma _{\mathrm{abs}}^{\mathrm{AdS_5}} $ and the Schwinger pair production rate$ \left|\mathfrak{b}^{\mathrm{AdS_5}}\right|^2 $ can be calculated by choosing the inner boundary condition$ {\cal D}_H^{(\mathrm{out})} = 0 $ and ($ c_3 = 0 $ ) and are given by$ \sigma _{{\rm{abs}}}^{{\rm{Ad}}{{\rm{S}}_{\rm{5}}}} = \left| {\frac{{{\cal D}_H^{({\rm{in}})}}}{{{\cal D}_B^{({\rm{in}})}}}} \right| = \frac{{{\rm{2}}T{L^{\rm{2}}}\left| \nu \right|\sinh \left( {2\pi \left| \Delta \right|} \right)}}{{{r_{\rm{o}}}{{\left| {H\left( {\nu ;\Delta ;\lambda } \right){{\left( {G_R^{{\rm{Ad}}{{\rm{S}}_{\rm{2}}}}} \right)}^{ - 1}} + H\left( { - \nu ;\Delta ;\lambda } \right)} \right|}^2}}}{\sigma _{{\rm{abs}}}}, $
(79) and
$ {\left| {{\mathfrak{b}^{{\text{Ad}}{{\text{S}}_{\text{5}}}}}} \right|^2} \!\!=\!\! \left| {\frac{{{\cal D}_H^{({\text{in}})}}}{{{\cal D}_B^{({\text{out}})}}}} \right| \!\!=\!\! \frac{{{\text{2}}T{L^{\text{2}}}\left| \nu \right|\sinh \left( {2\pi \left| \Delta \right|} \right)}}{{{r_{\rm{o}}}{{\left| {H\left( {\! -\! \nu ; \!-\! \Delta ;\lambda } \right)G_R^{{\text{Ad}}{{\text{S}}_{\text{2}}}} \!+\! H\left( {\nu ;\! -\! \Delta ;\lambda } \right)} \right|}^2}}}{\left| {{\mathfrak{b}^{{\text{Ad}}{{\text{S}}_{\text{2}}}}}} \right|^2}, $
(80) where
$ H\left( {x;y;z} \right) $ denotes a function$ H\left( {x;y;z} \right) \equiv {\left( { - 1} \right)^{2x}}{2^x}\frac{{\Gamma \left( {1 + {\rm{2}}x} \right)}}{{\Gamma \left( {\dfrac{1}{2} + x - y + z} \right)\Gamma \left( {\dfrac{1}{2} + x - y - z} \right)}}, $
(81) and
$ G_R^{\rm{AdS_2}} = {\left( { - 1} \right)^{4\nu - 1}}{2^{2\nu }}\frac{{\Gamma \left( {1 - 2\nu } \right)}}{{\Gamma \left( {1 + 2\nu } \right)}}\frac{{\Gamma \left( {\dfrac{1}{2} + \nu - {\rm i}{q_{{\rm{eff}}}}\ell } \right)}}{{\Gamma \left( {\dfrac{1}{2} - \nu - {\rm i}{q_{{\rm{eff}}}}\ell } \right)}}\frac{{\Gamma \left( {\dfrac{1}{2} + \nu - {\rm i}\dfrac{\omega }{{2\pi T}} + {\rm i}{q_{{\rm{eff}}}}\ell } \right)}}{{\Gamma \left( {\dfrac{1}{2} - \nu - {\rm i}\dfrac{\omega }{{2\pi T}}{\rm{ + }}{\rm i}{q_{{\rm{eff}}}}\ell } \right)}}, $ (82) which is exactly the retarded Green's function in Eq. (46) of the IR CFT in the near horizon, near extremal region. Furthermore,
$ {\sigma _{\mathrm{abs}}} $ and$ {\left| {{\mathfrak{b}^{{\text{Ad}}{{\text{S}}_{\text{2}}}}}} \right|^2} $ are exactly the absorption cross section ratio and the mean number of produced pairs of the corresponding IR CFT obtained from Eqs. (42) and (41). We find a relationship$\left|\mathfrak{b}^{\mathrm{AdS_5}}\right|^2 = -\sigma _{\mathrm{abs}}^{\mathrm{AdS_5}}\left(\left| \nu \right| \to - \left| \nu \right|, \left| \Delta \right| \to - \left| \Delta \right|\right), $
(83) which is similar to Eq. (43) except for a combined change in signs in both
$ |\nu| $ and$ |\Delta| $ .With Eq. (80) at hand we can easily investigate the relationship between the pair production rate in the near horizon and that for the whole spacetime of RN-AdS5. As shown in Fig. 1, we can see that the mean number of produced pairs for the whole spacetime is less than that from near horizon region. Moreover, with increasing charge of the scalar field, the corresponding ratio becomes smaller, which is consistent with previous assumptions stating that the Schwinger effect mainly occurs in the near horizon region.
Figure 1. (color online) Ratio of mean number of produced pairs for the whole spacetime to that in the near horizon region as a function of
$ \omega {L^2}/{r_{\rm{o}}} $ for different values of$ {q_{{\rm{eff}}}}\ell $ with$ T{L^{\text{2}}}/{{r_{\text{o}}}} = 0.001 $ ,$ \nu = 0.1{\rm i} $ , and$ \Delta = 0.1{\rm i} $ (left);$ T{L^{\text{2}}}/{{r_{\text{o}}}} = 0.001 $ ,$ \nu = 0.01{\rm i} $ , and$ \Delta = 0.01{\rm i} $ (middle);$ T{L^{\text{2}}}/{{r_{\text{o}}}} = 0.01 $ ,$ \nu = 0.1{\rm i} $ , and$ \Delta = 0.1i $ (right). -
To calculate the retarded Green's function, an ingoing boundary condition is required, namely
$ c_3 = 0 $ . Then, from Eqs. (72) and (73), the connection relations are$ {c_5} = {\left( { - 1} \right)^{ - 1 + \gamma - \beta - \lambda}}{c_4}\frac{{\Gamma \left( {2 - \gamma } \right)\Gamma \left( {\alpha - \beta } \right)}}{{\Gamma \left( {1 - \gamma + \alpha } \right)\Gamma \left( {1 - \beta } \right)}}, $
(84) $ {c_6} = {\left( { - 1} \right)^{ - 1 + \gamma - \alpha - \lambda}}{c_4}\frac{{\Gamma \left( {2 - \gamma } \right)\Gamma \left( {\beta - \alpha } \right)}}{{\Gamma \left( {1 - \gamma + \beta } \right)\Gamma \left( {1 - \alpha } \right)}}. $
(85) Substituting Eqs. (84) and (85) into Eq. (68) and taking
$ y \to \infty $ , namely the boundary of the AdS5 spacetime, one obtains$ \begin{aligned}[b] A(\tilde \omega ,\tilde k) = & \left( - 1 \right)^{{\rm i} q_{\rm{eff}}\ell - 2\lambda - 1 + \Delta }c_4\bigg( \frac{{\Gamma \left( {2 - \gamma } \right)\Gamma \left( {\alpha - \beta } \right)}}{{\Gamma \left( {1 - \gamma + \alpha } \right)\Gamma \left( {1 - \beta } \right)}}\frac{{\Gamma \left( {\gamma '} \right)\Gamma \left( {\alpha ' - \beta '} \right)}}{{\Gamma \left( {\alpha '} \right)\Gamma \left( {\gamma ' - \beta '} \right)}}+ \frac{{\Gamma \left( {2 - \gamma } \right)\Gamma \left( {\beta - \alpha } \right)}}{{\Gamma \left( {1 - \gamma + \beta } \right)\Gamma \left( {1 - \alpha } \right)}}\frac{{\Gamma \left( {2 - \gamma '} \right)\Gamma \left( {\alpha ' - \beta '} \right)}}{{\Gamma \left( {1 - \gamma ' + \alpha '} \right)\Gamma \left( {1 - \beta '} \right)}} \bigg),\\ \\ B(\tilde{\omega}, \tilde{k}) = & \left( - 1 \right)^{{\rm i} q_{\rm{eff}}\ell - 2\lambda - 1 - \Delta }c_4\bigg( \frac{{\Gamma \left( {2 - \gamma } \right)\Gamma \left( {\alpha - \beta } \right)}}{{\Gamma \left( {1 - \gamma + \alpha } \right)\Gamma \left( {1 - \beta } \right)}}\frac{{\Gamma \left( {\gamma '} \right)\Gamma \left( {\beta ' - \alpha '} \right)}}{{\Gamma \left( {\beta '} \right)\Gamma \left( {\gamma ' - \alpha '} \right)}}+ \frac{{\Gamma \left( {2 - \gamma } \right)\Gamma \left( {\beta - \alpha } \right)}}{{\Gamma \left( {1 - \gamma + \beta } \right)\Gamma \left( {1 - \alpha } \right)}}\frac{{\Gamma \left( {2 - \gamma '} \right)\Gamma \left( {\beta ' - \alpha '} \right)}}{{\Gamma \left( {1 - \gamma ' + \beta '} \right)\Gamma \left( {1 - \alpha '} \right)}} \bigg). \end{aligned} $ (86) Therefore, the retarded Green's function of the boundary CFT4 is given by
$ G_R^{\mathrm{AdS_5}}\sim\frac{B(\tilde{\omega}, \tilde{k})}{A(\tilde{\omega}, \tilde{k})} = {\left( { - 1} \right)^{ - 2\Delta }}\frac{{\dfrac{{\Gamma \left( {\alpha - \beta } \right)}}{{\Gamma \left( {1 - \gamma + \alpha } \right)\Gamma \left( {1 - \beta } \right)}}\dfrac{{\Gamma \left( {\gamma '} \right)\Gamma \left( {\beta ' - \alpha '} \right)}}{{\Gamma \left( {\beta '} \right)\Gamma \left( {\gamma ' - \alpha '} \right)}} + \dfrac{{\Gamma \left( {\beta - \alpha } \right)}}{{\Gamma \left( {1 - \gamma + \beta } \right)\Gamma \left( {1 - \alpha } \right)}}\dfrac{{\Gamma \left( {2 - \gamma '} \right)\Gamma \left( {\beta ' - \alpha '} \right)}}{{\Gamma \left( {1 - \gamma ' + \beta '} \right)\Gamma \left( {1 - \alpha '} \right)}}}}{{\dfrac{{\Gamma \left( {\alpha - \beta } \right)}}{{\Gamma \left( {1 - \gamma + \alpha } \right)\Gamma \left( {1 - \beta } \right)}}\dfrac{{\Gamma \left( {\gamma '} \right)\Gamma \left( {\alpha ' - \beta '} \right)}}{{\Gamma \left( {\alpha '} \right)\Gamma \left( {\gamma ' - \beta '} \right)}} + \dfrac{{\Gamma \left( {\beta - \alpha } \right)}}{{\Gamma \left( {1 - \gamma + \beta } \right)\Gamma \left( {1 - \alpha } \right)}}\dfrac{{\Gamma \left( {2 - \gamma '} \right)\Gamma \left( {\alpha ' - \beta '} \right)}}{{\Gamma \left( {1 - \gamma ' + \alpha '} \right)\Gamma \left( {1 - \beta '} \right)}}}}, $
(87) which is further simplified into
$ G_R^{\rm{AdS_5}} \sim {\left( { - 1} \right)^{ - 2\Delta }}\frac{{\Gamma \left( { - 2\Delta } \right)}}{{\Gamma \left( {2\Delta } \right)}}\frac{{H\left( {\nu ;\Delta ;\lambda} \right) + H\left( { - \nu ;\Delta ;\lambda} \right)G_R^{\rm{AdS_2}}}}{{H\left( {\nu ; - \Delta ;\lambda} \right) + H\left( { - \nu ; - \Delta ;\lambda} \right)G_R^{\rm{AdS_2}}}}. $
(88) -
From the AdS/CFT correspondence, the IR CFT1 in the near horizon, near extremal limit and the UV CFT4 at the asymptotic boundary of the RN-AdS5 black hole can be connected by the holographic RG flow [26, 27]. The CFT description of the Schwinger pair production in the IR region of charged black holes has been systematically studied in a series of previous works [3, 4, 6, 7]. Herein, we address the dual CFTs descriptions in the UV region and compare them with those in the IR region.
The IR CFT1 of the RN-AdS black hole is very similar to that of the RN black hole in an asymptotically flat spacetime, as CFT1 can be viewed as a chiral part of CFT2, which has the universal structures in its correlation functions. For instance, the absorption cross section of a scalar operator
$ {\cal O} $ in 2D CFT has the universal form$ \begin{aligned}[b] \sigma \sim & \frac{(2 \pi T_{\rm L})^{2h_{\rm L}-1}}{\Gamma(2 h_{\rm L})} \frac{(2 \pi T_{\rm R})^{2 h_{\rm R}-1}}{\Gamma(2 h_{\rm R})} \sinh\left( \frac{\omega_{\rm L} - q_{\rm L} \Omega_{\rm L}}{2 T_{\rm L}} + \frac{\omega_{\rm R} - q_{\rm R} \Omega_{\rm R}}{2 T_{\rm R}} \right) \\ & \times \left| \Gamma\left( h_{\rm L} + {\rm i} \frac{\omega_{\rm L} - q_{\rm L} \Omega_{\rm L}}{2 \pi T_{\rm L}} \right) \right|^2 \left| \Gamma\left( h_{\rm R} + {\rm i} \frac{\omega_{\rm R} - q_{\rm R} \Omega_{\rm R}}{2 \pi T_{\rm R}} \right) \right|^2, \end{aligned} $
(89) where
$ (h_{\rm L}, h_{\rm R}) $ ,$ (\omega_{\rm L}, \omega_{\rm R}) $ , and$ (q_{\rm L}, q_{\rm R}) $ are the left- and right-hand conformal weights, excited energies, charges associated with operator$ {\cal O} $ , respectively, while$ (T_{\rm L}, T_{\rm R}) $ and$ (\Omega_{\rm L}, \Omega_{\rm R}) $ are the temperatures and chemical potentials of the corresponding left- and right-hand sectors of the 2D CFT. Further identifying the variations in the black hole area entropy with those of the CFT microscopic entropy, namely$ \delta S_{\rm BH} = \delta S_{\rm CFT} $ , one derives$ \frac{\delta M}{T_H} - \frac{\Omega_H \delta Q}{T_H} = \frac{\omega_{\rm L} - q_{\rm L} \Omega_{\rm L}}{T_{\rm L}} + \frac{\omega_{\rm R} - q_{\rm R} \Omega_{\rm R}}{T_{\rm R}}, $
(90) where the left hand side of Eq. (90) is calculated with coordinate (14), for which
$ \delta M = \xi_{\mathrm o}w $ ,$ \delta Q = q $ ,$ T_H = \tilde{T}_{ n} $ , and$ \Omega_H = 2\mu\ell^2/r_{\mathrm o} $ , and thus, it is equal to$ w/T_{ n}-2\pi q_{\mathrm{eff}}\ell $ . Moreover, the violation of the BF bound in AdS2 makes the conformal weights of the scalar operator$ {\cal O} $ dual to$ \phi $ a complex, which can be chosen as$ h_{\rm L} = h_{\rm R} = \dfrac 1 2+{\rm i}|\nu| $ , even without further knowledge about the central charge and$ (T_{\rm L}, T_{\rm R}) $ of the IR CFT dual to the near extremal RN-AdS5 black hole. One can also see that the absorption cross section ratio (42) in the AdS2 spacetime has the form of Eq. (89) up to some coefficients, depending on the mass and charge of the scalar field.In contrast, the absorption cross section and retarded Green's functions in a general 4D finite temperature CFT cannot be as easily calculated in momentum space as in the 2D CFT. Thus, it is not straightforward to compare the calculations between the bulk gravity and the boundary CFT sides. Nevertheless, from Eqs. (79) and (80), both the absorption cross section ratio
$ \sigma _{\mathrm{abs}}^{\mathrm{AdS_5}} $ and the Schwinger mean number of produced pairs$ \left|\mathfrak{b}^{\mathrm{AdS_5}}\right|^2 $ calculated from the bulk near extremal RN-AdS5 black hole have a simple proportional relation with their counterparts in the near horizon region. Moreover, the violation of the BF bound (77) in AdS5 spacetime indicates the complex conformal weights$ \bar{\Delta} = 2+2{\rm i}|\Delta| $ of the scalar operator$ \bar{{\cal O}} $ in the UV 4D CFT at the asymptotic spatial boundary of the RN-AdS5 black hole, which also indicates that, to have pair production in the full bulk spacetime, the corresponding operators in the UV CFT should be unstable. Interestingly, Eq. (83) shows that under the interchange between the roles of source and operator both in the IR and UV CFTs at the same time, namely$ h_{\rm L,R} = \dfrac12+ $ $ {\rm i}|\nu| \to \dfrac12-{\rm i}|\nu| $ and$ \bar{\Delta} = 2+2{\rm i}|\Delta|\to 2-2{\rm i}|\Delta| $ , the full absorption cross section ratio$ \sigma _{\mathrm{abs}}^{\mathrm{AdS_5}} $ and the Schwinger pair production rate$ \left|\mathfrak{b}^{\mathrm{AdS_5}}\right|^2 $ are interchanged with each other only up to a minus sign. Note that both the charge and the mass of the scalar particle contribute to the conformal weights$ h_{\rm L,R} $ of the scalar operator in the dual IR CFT; however, only the mass contributes to the conformal weight$ \bar{\Delta} $ of the scalar operator in the dual UV CFT. Actually, it can be seen from the expressions of the conformal weights that the non-zero charge and mass for the scalar field are crucial for the violation of the BF bound in the corresponding AdS spacetimes and hence guarantee the existence of the Schwinger pair production. However, when the charge of the particle is zero, there will be no Schwinger effect, except for an exponentially suppressed Hawking radiation in near extremal black holes.
Pair production in Reissner-Nordström-Anti de Sitter black holes
- Received Date: 2021-02-01
- Available Online: 2021-06-15
Abstract: We studied the pair production of charged scalar particles of a five-dimensional near extremal Reissner-Nordström-Anti de Sitter (RN-AdS5) black hole. The pair production rate and the absorption cross section ratio in full spacetime are obtained and are shown to have a concise relation with their counterparts in the near horizon region. In addition, the holographic descriptions of the pair production, both in the IR CFT in the near horizon region and the UV CFT at the asymptotic spatial boundary of the RN-AdS5 black hole, are analyzed in the AdS2/CFT1 and AdS5/CFT4 correspondences, respectively. This work gives a complete description of scalar pair production in a near extremal RN-AdS5 black hole.