Exact low-temperature Green’s functions in AdS/CFT: From Heun to confluent Heun

Paolo Arnaudo, Benjamin Withers · 2024

Read Exact low-temperature Green’s functions in AdS/CFT: From Heun to confluent Heun by Paolo Arnaudo, Benjamin Withers (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-01923.txt.

Introduction

Paolo Arnaudo and Benjamin Withers compute exact response functions in 2024 for a holographic model of a superconductor. Their setting pairs a quantum field theory with a charged black hole in a curved space, a duality used to study strong interactions. At low temperature the math hinges on a special differential equation family.

Using new results from a related theory, they solve the connection problem in a controlled double limit where frequency and temperature fall together. From the response function they extract the critical temperature for the holographic superconductor. They also derive dispersion formulas for gapped and gapless low-temperature modes. The exact results match numerical work precisely.

Catalog record

CREATORPaolo Arnaudo, Benjamin Withers
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE47 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-01923

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Exact low-temperature Green’s functions in AdS/CFT: From Heun to confluent Heun Paolo Arnaudo, Benjamin Withers Abstract We obtain exact expressions for correlation functions of charged scalar operators at finite density and low temperature in CFT 4 dual to the RN-AdS 5 black brane. We use recent developments in the Heun connection problem in black hole perturbation theory arising from Liouville CFT and the AGT correspondence. The connection problem is solved perturbatively in an instanton counting parameter, which is controlled in a double-scaling limit where ω , T → 0 \omega,T\to 0 holding ω / T \omega/T fixed. This provides analytic control over the emergence of the zero temperature branch cut as a confluent limit of the Heun equation. From the Green’s function we extract analytic results for the critical temperature of the holographic superconductor, as well as dispersion relations for both gapped and gapless low temperature quasinormal modes. We demonstrate precise agreement with numerics. I Introduction and results The real-time linear response of strongly coupled quantum field theories (QFTs) at finite temperature and density is a challenging domain where the techniques of holographic duality [ 1 , 2 ] can be exploited to perform first-principles computations. In this context, retarded correlators are of considerable interest in the context of condensed matter physics [ 3 , 4 ] , nuclear collisions [ 5 , 6 , 7 ] , and neutron stars [ 8 , 9 ] . In this work we analytically compute retarded correlation functions at finite density and low temperature by exploiting recent developments connecting black hole perturbation theory to Seiberg-Witten theory [ 10 ] , and subsequently the connection problem for the Heun equation [ 11 , 12 , 13 , 14 ] , which governs the linear response problem in holography. We gain unprecedented insight into the analytic structure of low-temperature correlators, including an exact description of the coalescence of poles into branch cuts at zero temperature, as well as analytic control over pole locations and the superconducting phase transition. Other recent works in the holographic context that have utilised such techniques are [ 15 , 16 ] . The focus of this work are retarded Green’s functions of scalar operators of conformal dimension \Delta , \displaystyle G^{R}(x,y)=-i\theta(x^{0}-y^{0})\left<[{\cal O}_{\Phi}(x),{\cal O}_{\Phi}(y)]\right>, (1) in momentum space, \tilde{G}^{R}(p)=\int d^{4}xG^{R}(0,x)e^{-ipx} with p^{\mu}=(\omega,\vec{k}) . We focus on the 3+1 -dimensional QFT dual to the Reissner-Nordström (RN)-AdS 5 black brane at chemical potential \mu , as a solution of the classical equations of motion of the action, S=\int\sqrt{-g}d^{5}x(R-F^{2}+12) , where F=dA , given by the following spacetime metric, \displaystyle ds^{2} \displaystyle= \displaystyle-f(r)dt^{2}+\frac{dr^{2}}{f(r)}+r^{2}d\vec{x}^{2}, (2) \displaystyle f(r) \displaystyle= \displaystyle r^{2}-\frac{1+\frac{4\mu^{2}}{3}}{r^{2}}+\frac{4\mu^{2}}{3r^{4}} (3) and gauge field, A=\mu\left(1-\frac{1}{r^{2}}\right)dt. (4) Here the AdS radius is one, the event horizon is located at r=1 and the temperature T=\frac{3-2\mu^{2}}{3\pi} . Figure 1: Comparison of analytic and numerical results for the retarded Green’s function. Solid curves show the analytic expressions ( 5 ) while dots are numerics. We take \Delta=5/2 , \hat{k}=0 and \nu^{2}=-0.07 at \hat{T}=4\times 10^{-4} , where \hat{T}_{c}=3.4\times 10^{-4} . The visible peak is the imprint of the superconductor zero mode since we are near \hat{T}_{c} . Here N_{\rho}=400 . Upper panel: Along the real \tilde{\omega} axis. Lower panel: Along the imaginary \tilde{\omega} axis. To compute ( 1 ) in this state we consider linearised charged scalar field perturbations \Phi on this background, dual to the scalar operator {\cal O}_{\Phi} . \Phi obeys D^{2}\Phi=m_{\Phi}^{2}\Phi where D\equiv\nabla-ieA and m_{\Phi}^{2}=\Delta(\Delta-4) . We restrict to \Delta>2 and present our method at generic values of \Delta . Our results are as follows. We denote variables made dimensionless with \mu using hats, i.e. \hat{T}=\frac{T}{\mu} , \hat{k}=\frac{k}{\mu} , \hat{\omega}=\frac{\omega}{\mu} and we consider the double scaling regime \hat{T}\to 0 and \hat{\omega}\to 0 holding \tilde{\omega}=\frac{\omega}{T} fixed. We compute \tilde{G}^{R}(\tilde{\omega},\hat{k}) , which takes the following form \tilde{G}^{R}(\tilde{\omega},\hat{k})=N\frac{\mathcal{C}_{4-\Delta,\nu}(\tilde{\omega},\hat{k})-\mathcal{C}_{4-\Delta,-\nu}(\tilde{\omega},\hat{k})\,{\cal G}(\tilde{\omega},\hat{k})}{\mathcal{C}_{\Delta,\nu}(\tilde{\omega},\hat{k})-\mathcal{C}_{\Delta,-\nu}(\tilde{\omega},\hat{k})\,{\cal G}(\tilde{\omega},\hat{k})}, (5) where the various terms appearing N,\mathcal{C}_{\Delta,\nu},\mathcal{G} are determined analytically as a systematic expansion in \hat{T} within the double scaling regime, \mathcal{C}_{\Delta,\nu}(\tilde{\omega},\hat{k})=\mathcal{C}^{(0)}_{\Delta,\nu}(\hat{k})+\mathcal{C}^{(1)}_{\Delta,\nu}(\tilde{\omega},\hat{k})\,\hat{T}+\mathcal{O}\left(\hat{T}^{2}\right), (6) \displaystyle{\cal G}(\tilde{\omega},\hat{k}) \displaystyle= \displaystyle{\cal G}^{(0)}(\tilde{\omega},\hat{k})\hat{T}^{2\nu}+{\cal G}^{(1,+)}(\tilde{\omega},\hat{k})\hat{T}^{2\nu+1}\log\hat{T} (7) \displaystyle+{\cal G}^{(1,-)}(\tilde{\omega},\hat{k})\hat{T}^{2\nu+1}+\mathcal{O}\left(\hat{T}^{2\nu+2}\log{\hat{T}}\right) N=2(\Delta-2)\,3^{\Delta-2}+\mathcal{O}(\hat{T}). (8) We present explicit expressions for the contributions appearing in ( 5 ): \mathcal{C}^{(0)}_{\Delta,\nu}(\hat{k}) ( 54 ), \mathcal{C}^{(1)}_{\Delta,\nu}(\tilde{\omega},\hat{k}) ( 55 ), {\cal G}^{(0)}(\tilde{\omega},\hat{k}) ( 9 ), {\cal G}^{(1,\pm)}(\tilde{\omega},\hat{k}) ( 56 )-( 57 ). One may straightforwardly extend this series to higher orders using the methodology presented. In particular note that {\cal G}^{(0)}\hat{T}^{2\nu} can be interpreted as a finite- \hat{T} AdS 2 Green’s function, where \displaystyle{\cal G}^{(0)}(\tilde{\omega},\hat{k})=\left(\frac{2\pi^{2}}{27}\right)^{\nu}\frac{\Gamma(-2\nu)^{2}}{\Gamma(2\nu)^{2}} (9) \displaystyle\qquad\qquad\times\frac{\Gamma\left(\frac{1}{2}-\frac{ie}{2\sqrt{6}}+\nu\right)}{\Gamma\left(\frac{1}{2}-\frac{ie}{2\sqrt{6}}-\nu\right)}\frac{\Gamma\left(\frac{1}{2}+\nu+\frac{ie}{2\sqrt{6}}-\frac{i\tilde{\omega}}{2\pi}\right)}{\Gamma\left(\frac{1}{2}-\nu+\frac{ie}{2\sqrt{6}}-\frac{i\tilde{\omega}}{2\pi}\right)}, and where the parameter \nu\equiv\frac{1}{2}\,\sqrt{1+\frac{\Delta(\Delta-4)}{3}+\frac{\hat{k}^{2}}{2}-\frac{e^{2}}{6}} (10) labels representations of the near-horizon AdS 2 theory at \hat{T}=0 with conformal dimensions \delta_{\pm}=\frac{1}{2}\pm\nu [ 17 ] . In particular, \tilde{G}^{R} is invariant under \nu\to-\nu . Some parts of this analytic structure were seen in low \hat{\omega} perturbation theory near extremality [ 17 , 18 , 19 , 20 ] . In [ 17 , 18 , 19 ] the structure seen in ( 9 ) was obtained as a correlator for a black hole inside AdS 2 . In [ 20 ] the leading matching coefficients for a small- \hat{\omega} expansion were obtained using hypergeometric connection formulae which arise at extremality, and we find agreement in that limit. 1 1 1 We note q_{\text{there}}=e_{\text{here}}/2 , T_{\text{there}}=\sqrt{3/2}\hat{T}_{\text{here}} and {\cal G}_{k,\text{there}}^{(T)}=2^{2\nu}3^{4\nu}\Gamma(2\nu){\cal G}^{(0)}_{\text{here}}/\Gamma(-2\nu) . See also [ 21 , 22 , 23 , 24 , 25 , 26 ] for other related low \hat{\omega},\hat{T} matching computations involving AdS 2 critical points. Here we emphasise that the Heun connection problem is naturally controlled by low \hat{T} at fixed \tilde{\omega}=\omega/T=\mathcal{O}(1) , which provides a systematic treatment of the correlators – indeed, \tilde{\omega} appears non-trivially inside each of {\cal G}^{(0)}(\tilde{\omega},\hat{k}) , \mathcal{C}^{(1)}_{\Delta,\nu}(\tilde{\omega},\hat{k}) , {\cal G}^{(1,\pm)}(\tilde{\omega},\hat{k}) . Figure 2: The analytic structure of \tilde{G}^{R} at low {\hat{T}}\sim\hat{\omega} in the \hat{\omega} plane, described using closed-form solutions to the Heun connection problem. The analytic expression for \tilde{G}^{R} ( 5 ) captures the process where the line of poles ( 12 ) coalesces in to a branch cut at \hat{T}=0 , as a confluent limit of the Heun equation via ( 11 ). Optionally, by tuning model parameters so that the holographic superconductor critical temperature \hat{T}_{c}\ll 1 , then ( 5 ) also captures the holographic superconductor zero mode ( 15 ),( 36 ) (red asterisk) and \hat{T}_{c} itself. The closed-form Green’s function ( 5 ) is the full RN-AdS 5 correlator – rather than an AdS 2 near-horizon result – and it displays an intricate analytic structure for which we find excellent agreement with low-temperature numerics, as shown in figure 1 . Moreover, we note that taking {\hat{T}}\to 0 holding \hat{\omega} fixed one has \hat{T}^{2\nu}\frac{\Gamma\left(\frac{1}{2}+\nu+\frac{ie}{2\sqrt{6}}-\frac{i\hat{\omega}}{2\pi\hat{T}}\right)}{\Gamma\left(\frac{1}{2}-\nu+\frac{ie}{2\sqrt{6}}-\frac{i\hat{\omega}}{2\pi\hat{T}}\right)}\to\left(-\frac{i\hat{\omega}}{2\pi}\right)^{2\nu}, (11) where we used the Stirling approximation \Gamma(z+\alpha)/\Gamma(z)\sim z^{\alpha} for z\to\infty and \alpha\in\mathbb{C} , and then took the \hat{T}\to 0 limit. This analytically describes a line of poles coalescing into a branch point at \hat{\omega}=0 , capturing the process previously observed numerically in [ 27 , 28 , 29 ] in the context of RN-AdS 4 . This is illustrated in figure 2 . When \mathfrak{Re}(\nu)>0 the line of poles in question can be determined analytically. Because of the presence of the term \hat{T}^{2\nu} , these poles can be read off from the poles of the \Gamma -function in the numerator of \mathcal{G}^{(0)} , \displaystyle\tilde{\omega} \displaystyle= \displaystyle\tilde{\omega}_{0}+\tilde{\omega}_{1}\hat{T}+\tilde{\omega}_{2\nu}\hat{T}^{2\nu}+O(\hat{T}^{2},\hat{T}^{4\nu}), (12) \displaystyle\tilde{\omega}_{0} \displaystyle= \displaystyle\frac{\pi e}{\sqrt{6}}-i\pi(2\nu+2n+1) (13) where n\in\mathbb{Z}_{\geq

Read the complete text