Two-point tree-level string amplitudes as AdS transition amplitudes

Pongwit Srisangyingcharoen, Jongruk Pukdee · 2025

Read Two-point tree-level string amplitudes as AdS transition amplitudes by Pongwit Srisangyingcharoen, Jongruk Pukdee (2025) 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-2505-07358.txt.

Introduction

Pongwit Srisangyingcharoen and Jongruk Pukdee compute basic string amplitudes in a 2025 theory paper. They work out the two-point scattering strength for both open and closed strings at the simplest level. In a strong-coupling style limit, the results take a shape that mirrors how a field crosses a curved anti-de Sitter space from wall to wall.

Open and closed strings give matching expressions, which is striking. Yet the effective coupling for each is set by a different kind of worldsheet curvature, one based on path length and one on surface bending. Readers get a hint that open-closed duality may carry a geometric side worth probing.

Catalog record

CREATORPongwit Srisangyingcharoen, Jongruk Pukdee
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE25 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-07358

Browse the full library · About the Tower of Babel

Full text

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

Two-point tree-level string amplitudes as AdS transition amplitudes Pongwit Srisangyingcharoen, Jongruk Pukdee Abstract We compute the two-point open string and closed string amplitudes at tree level and show that, in a ’t Hooft-like limit, they take a form structurally analogous to boundary-to-boundary transition amplitudes of a scalar field in Euclidean AdS space. Interestingly, while both amplitudes yield equivalent expressions, the associated ’t Hooft couplings are defined by different worldsheet curvatures—geodesic for the disk and Gaussian for the sphere—suggesting a possible geometric aspect of open/closed string duality. 1 Introduction An essential framework for exploring the interplay between gravity and quantum field theory is the AdS/CFT correspondence [ 1 ] . The idea that quantum gravity may be fundamentally holographic was first inspired by the thermodynamic properties of black holes, particularly the Bekenstein–Hawking entropy formula [ 2 , 3 ] . The AdS/CFT correspondence offers the first concrete realization of this holographic principle within string theory [ 1 , 4 , 5 ] . Beyond its conceptual significance, the correspondence has become a powerful computational tool for probing strongly coupled, non-perturbative regimes of quantum field theories. See [ 6 , 7 , 8 ] for a comprehensive review. A particularly intriguing perspective on this duality was presented by Gopakumar [ 9 ] . In his work, he showed that one-loop planar correlation functions in free field theories can be reorganized into a form that mimics field amplitudes in AdS space. This was accomplished by expressing Feynman diagrams as integrals over Schwinger parameters, which naturally matched with the bulk-to-boundary propagators of AdS scalar fields. Motivated by this insight, we investigate whether similar structures arise directly from open string amplitudes in flat space, without appealing to a field theory limit. In particular, we analyze the two-point open string amplitude at tree level and study whether, in a suitable limit, it can be cast into a form structurally resembling an AdS boundary-to-boundary transition amplitude. Our goal is to understand if and how open string correlators, under appropriate parametrization and scaling, reveal an intrinsic connection to propagators in AdS space—potentially offering a new perspective on the emergence of holographic dualities. 2 AdS Scalar propagator and transition amplitude In this section, we would like to exhibit the bulk-to-boundary propagators in \text{AdS}_{d+1} in a Schwinger representation for a massive scalar field. It is convenient to work in the Poincaré coordinates of Euclidean \text{AdS}_{d+1} , ds^{2}=\frac{d\vec{z}^{2}+dz_{0}^{2}}{z_{0}^{2}}. (1) Let us consider the Euclidean action S[\phi]=\frac{1}{2}\int d^{d}zdz_{0}\sqrt{g}\left(g^{MN}\partial_{M}\phi(z)\partial_{N}\phi(z)+m^{2}\phi(z)\right) (2) where \phi(z)=\phi(z_{0},\vec{z}) . A bulk-to-boundary propagator K(z_{0},\vec{z};\vec{z}^{\prime}) can be obtained from the wave equation \left(z_{0}^{d+1}\frac{\partial}{\partial z_{0}}\left(z_{0}^{-d+1}\frac{\partial}{\partial z_{0}}\right)+z_{0}^{2}\Box-m^{2}\right)K=0. (3) where \Box is d -dimensional Laplacian operator in direction \vec{z} . It is well-known that the expression for bulk-to-boundary propagator takes the form K(z_{0},\vec{z};\vec{z}^{\prime})=\frac{\Gamma(\Delta)}{\pi^{d/2}\Gamma(\Delta-\frac{d}{2})}\left(\frac{z_{0}}{z_{0}^{2}+(\vec{z}-\vec{z}^{\prime})^{2}}\right)^{\Delta} (4) where \Delta=\frac{1}{2}\left(d+\sqrt{d^{2}+4m^{2}}\right) [ 4 ] . The AdS bulk field \phi(z) is related to the boundary data by \phi(z_{0},\vec{z})=\int d^{d}\vec{x}\ K(z_{0},\vec{z};\vec{x})\phi_{0}(\vec{x}). (5) Using a Schwinger parametrization, one obtains K(z_{0},\vec{z};\vec{z}^{\prime})=\frac{1}{z_{0}^{\Delta}\pi^{d/2}\Gamma(\Delta-\frac{d}{2})}\int_{0}^{\infty}d\rho\ \rho^{\Delta-1}e^{-\rho}e^{-\frac{(\vec{z}-\vec{z}^{\prime})^{2}}{z_{0}^{2}}\rho} (6) by which we can utilize the heat kernel, \langle x|e^{\tau\Box}|y\rangle=\frac{1}{(4\pi\tau)^{d/2}}e^{-(x-y)^{2}/4\tau} (7) to rewrite K(t,\vec{z};\vec{z}^{\prime})=\frac{t^{(d-\Delta)/2}}{\Gamma(\Delta-\frac{d}{2})}\int_{0}^{\infty}d\rho\ \rho^{\Delta-\frac{d}{2}-1}e^{-\rho}\langle\vec{z}|e^{\frac{t}{4\rho}\Box}|\vec{z}^{\prime}\rangle (8) where t=z_{0}^{2} . We would write \vec{z} as z for later convenience. We can then construct a boundary-to-boundary propagator through \displaystyle\Gamma(x,y)= \displaystyle\frac{1}{2}\int_{0}^{\infty}\frac{dt}{t^{\frac{d}{2}+1}}\int d^{d}z\ K(t,z;y)K(t,z;x) \displaystyle= \displaystyle\frac{1}{2(\Gamma(\Delta-d/2))^{2}}\int_{0}^{\infty}\frac{dt}{t^{\Delta+1-\frac{d}{2}}}\int d^{d}z\int_{0}^{\infty}d\rho_{1}\int_{0}^{\infty}d\rho_{2} \displaystyle\times\ (\rho_{1}\rho_{2})^{\Delta-\frac{d}{2}-1}e^{-(\rho_{1}+\rho_{2})}\langle y|e^{\frac{t}{4\rho_{2}}\Box}|z\rangle\langle z|e^{\frac{t}{4\rho_{1}}\Box}|x\rangle. (9) While equation ( 9 ) describes the standard boundary-to-boundary propagator constructed from two identical bulk-to-boundary propagators of weight \Delta , one may consider a generalized form involving distinct weights \Delta_{1} and \Delta_{2} . This leads to \displaystyle\Gamma(x,\Delta_{1};y,\Delta_{2})=\frac{1}{2\Gamma(\Delta_{1}-\frac{d}{2})\Gamma(\Delta_{2}-\frac{d}{2})}\int_{0}^{\infty}\frac{dt}{t^{1+(\Delta_{1}+\Delta_{2}-d)/2}}\int d^{d}z \displaystyle\times\int_{0}^{\infty}d\rho_{1}\int_{0}^{\infty}d\rho_{2}\ \rho_{1}^{\frac{\Delta_{1}-d}{2}-1}\rho_{2}^{\frac{\Delta_{2}-d}{2}-1}e^{-(\rho_{1}+\rho_{2})}\langle y|e^{\frac{t}{4\rho_{2}}\Box}|z\rangle\langle z|e^{\frac{t}{4\rho_{1}}\Box}|x\rangle (10) which can be interpreted as a transition amplitude between scalar fields of different weights. For future reference, let us denote \displaystyle 2\Gamma(\Delta_{1}-d/2) \displaystyle\Gamma(\Delta_{2}-d/2)\Gamma(x,\Delta_{1};y,\Delta_{2}) \displaystyle=\Gamma_{1}(x,\Delta_{1};y,\Delta_{2})+\Gamma_{2}(x,\Delta_{1};y,\Delta_{2}) (11) where \Gamma_{1} and \Gamma_{2} refer to ( 10 ) evaluated in different integration domains: \rho_{1}<\rho_{2} for \Gamma_{1} and \rho_{1}>\rho_{2} for \Gamma_{2} by which we refer to this integration region as \mathcal{I}_{1} and \mathcal{I}_{2} respectively. Furthermore, we would like to introduce t -dependent transition amplitude as \displaystyle\Gamma_{i}(x,\Delta_{1};y,\Delta_{2};t)= \displaystyle\int d^{d}z\int_{\mathcal{I}_{i}}d\rho_{1}d\rho_{2}\ \rho_{1}^{\frac{\Delta_{1}-d}{2}-1}\rho_{2}^{\frac{\Delta_{2}-d}{2}-1} \displaystyle\times e^{-(\rho_{1}+\rho_{2})}\langle y|e^{\frac{t}{4\rho_{2}}\Box}|z\rangle\langle z|e^{\frac{t}{4\rho_{1}}\Box}|x\rangle. (12) Intuitively, this function could be interpreted as a partial transition amplitude which is in the same sense as color-ordered amplitude in QFT or in the context of open string amplitude, where the vertex operators of the states are inserted along the worldsheet boundary in an ordering way. This can be illustrated in figure( 1 ) as a propagation of the scalar field \phi_{\Delta_{1}}(\vec{x},0) from the point on the boundary (parameterized by \rho_{1} ) to the point (\vec{z},t_{0}) in the bulk. At this point, the field transitions into the field with different weight \Delta_{2} and then propagates to the boundary point (\vec{y},0) parameterized by \rho_{2} . 0 \rho t • \rho_{1} • \rho_{2} • - t_{0} \phi_{\Delta_{1}} \phi_{\Delta_{2}} Figure 1: A propagation of the scalar field in configuration space from the boundary to the boundary through an interaction point located in the bulk. 3 Two-point Open String Amplitudes In string theory, amplitudes are computed as correlation functions of vertex operators on Riemann surfaces. For open strings, these operators are inserted along the boundary of a disk, which is typically mapped conformally to the upper half-plane. It was widely believed that two-point string amplitudes vanish due to residual worldsheet symmetries after fixing two insertion points which have infinite volume. Until quite recently, it is pointed out that the resulting divergence cancels the \delta(0) arising from momentum conservation in the path integral formulation [ 10 ] . Since then, few follow-up works have explored this issue using alternative approaches [ 11 , 12 , 13 , 14 ] . In this paper, we would like to compute a two-point open string amplitude with the aim of recasting its form to resemble that of a boundary-to-boundary AdS transition amplitude ( 10 ). The motivation is to explore manifestations of the AdS/CFT correspondence, under the expectation that the string amplitude—interpreted from the CFT side—would exhibit a structure analogous to the AdS amplitude. An expression for the two-point open string amplitude at tree level is given by A(k_{1},k_{2})=\frac{1}{Z}\int DXDg\ \mathcal{V}_{1}(k_{1})\mathcal{V}_{2}(k_{2})e^{-S_{P}[X,g]} (13) where the Polyakov action is S_{P}[X,g]=\frac{1}{4\pi\alpha^{\prime}}\int_{\Sigma}d^{2}\sigma\sqrt{g}g^{ab}\partial_{a}X^{\mu}\partial_{b}X^{\nu}\eta_{\mu\nu}. (14) \mathcal{V}_{i}(k_{i}) is an integrated vertex operator taking the form \mathcal{V}_{i}(k_{i})=g_{o}\int_{\partial\Sigma}dx_{i}\ V_{i}(x_{i},k_{i}) (15) with an open-string coupling constant g_{o} . The denominator Z is a partition function given by \int DXDg\ e^{-S_{P}[x,g]}. (16) Using integration by parts, the action becomes \frac{1}{4\pi\alpha^{\prime}}\int_{\Sigma}d^{2}\sigma\left(\sqrt{g}X^{\mu}\nabla^{2}X_{\mu}-\partial_{a}\left(\sqrt{g}g^{ab}X^{\mu}\partial_{b}X_{\mu}\right)\right) (17) where \nabla^{2}=\frac{1}{\sqrt{g}}\partial_{a}\left(\sqrt{g}g^{ab}\partial_{b}\right) (18) is the Laplace–Beltrami operator. The second term represents a boundary contribution, which vanishes under Neumann boundary conditions, i.e., \frac{\partial X}{\partial n}\biggr\rvert_{\partial\Sigma}=0 . For simplicity, we fix the worldsheet metric to g^{ab}=\delta^{ab} , which leaves the residual gauge symmetry group as PSL (2,\mathbb{R}) . To simplify the computation, we restrict our attention to amplitudes involving the simplest form of vertex operators: V_{j}(x_{j},k_{j})=e^{ik_{j}\cdot X(x_{j})}. (19) with a Euclidean on-shell condition k_{i}^{2}=-\frac{1}{\alpha^{\prime}} (We use mostly minus Minkowski signature, accordingly, the Euclidean mass-shell condition reads k^{2}=-k_{M}^{2}=m^{2} upon a Wick rotation). Expanding X(\sigma)=\bar{X}(\sigma)+x where x is a zero mode regarding the Laplacian operator, the measure reads D\bar{X}d^{d}x . Obviously the integration from the zero mode x provides a momentum conservation, (2\pi)^{d}\delta^{d}(k_{1}+k_{2}). (20) What remains is a Gaussian integral, \displaystyle\int D\bar{X}\exp\Bigg( \displaystyle\int_{\Sigma}d^{2}\sigma\Bigg[\frac{\sqrt{g}}{4\pi\alpha^{\prime}}X^{\mu}\nabla^{2}X_{\mu} \displaystyle\quad+i\sum_{i=1}^{2}\delta^{2}(\sigma-x_{i})k_{i}\cdot X(\sigma)\Bigg]\Bigg). (21) After performing a Gaussian integration, one would obtain \mathcal{N}\exp{\left(\pi\alpha^{\prime}k_{1}\cdot k_{2}G(x_{1}-x_{2})\right)} (22) where G(x_{1},x_{2}) is a Green’s function satisfying \nabla^{2}G(x,y)=\frac{1}{\sqrt{g}}\delta^{2}(x-y)-\frac{1}{\text{Vol}} (23) with the Neumann boundary condition \frac{\partial G(x,y)}{\partial n_{x}}\Biggr|_{\partial\Sigma}=c\quad\text{(constant)}. (24) Note that the factor \mathcal{N}

Read the complete text