Witten–Sakai–Sugimoto Model Overview

• The Witten–Sakai–Sugimoto model is a top-down holographic realization of low-energy QCD using a confining D4-brane background and smoothly connecting D8-branes to encode chiral symmetry breaking.
• It derives mesons, glueballs, and baryons from a common higher-dimensional framework, yielding quantitative predictions for hadronic masses and decay widths.
• The model integrates anomaly physics, dense matter phenomena, and thermal transitions while acknowledging limitations from large-Nc, large-λ approximations and Kaluza–Klein modes.

The Witten–Sakai–Sugimoto (WSS) model is a top-down holographic model of low-energy QCD derived from type-IIA string theory, built from Witten’s confining D4-brane background together with Sakai–Sugimoto probe D8/$\overline{\mathrm{D8}}$-branes. In this construction, confinement is encoded by the cigar geometry of the compactified D4 background, while spontaneous chiral symmetry breaking is realized geometrically by the joining of the D8 and $\overline{\mathrm{D8}}$ branes in the bulk. The model is formulated in the large-$N_c$, large-$\lambda$ regime, and—once the Kaluza–Klein scale $M_{\rm KK}$ and the ’t Hooft coupling $\lambda$ are fixed—generates mesons, glueballs, baryons, anomaly physics, and finite-density phases from a common higher-dimensional setup rather than from phenomenological insertion of separate hadronic sectors (Rebhan, 2014).

1. String-theoretic construction and holographic regime

Witten’s original construction starts from $N_c$ D4-branes with one spatial direction compactified on a circle with antiperiodic boundary conditions for fermions. This compactification breaks supersymmetry, removes conformal symmetry, and introduces a finite radius whose inverse defines the Kaluza–Klein scale $M_{\rm KK}$. At energies below that scale the model is approximately four-dimensional and confining, while the supergravity description is reliable for large ’t Hooft coupling $\lambda=g_{\mathrm{YM}}^2N_c$ (Parganlija, 2015).

Flavor is added by introducing $N_f\ll N_c$ probe D8- and $\overline{\mathrm{D8}}$0-branes. In the antipodal embedding, the two stacks join smoothly in the bulk, geometrically realizing

$\overline{\mathrm{D8}}$1

This identifies the connected brane configuration with spontaneous chiral symmetry breaking. Mesons arise from gauge-field fluctuations on the flavor branes, whereas glueballs arise from normalizable fluctuations of bulk supergravity fields, including metric, dilaton, and Ramond–Ramond sectors (Brünner et al., 2015).

The standard phenomenological matching fixes

$\overline{\mathrm{D8}}$2

by the $\overline{\mathrm{D8}}$3-meson mass and the pion decay constant. A frequently used alternative is

$\overline{\mathrm{D8}}$4

motivated by matching the large-$\overline{\mathrm{D8}}$5 string tension. This leaves the model “almost parameter-free” in the sense that, once $\overline{\mathrm{D8}}$6 is chosen, essentially the only dimensionless parameter is $\overline{\mathrm{D8}}$7 (Rebhan, 2016).

The model is not exact QCD. The review literature emphasizes several structural limitations: it is a large-$\overline{\mathrm{D8}}$8, large-$\overline{\mathrm{D8}}$9, probe-flavor construction; Kaluza–Klein modes cannot be decoupled while remaining in the controlled supergravity regime; and many quantitative comparisons rely on extrapolation from the holographic regime to $N_c$0 and moderate coupling (Rebhan, 2014).

2. Flavor-brane effective action, mesons, and anomaly structure

The flavor dynamics is governed by the D8-brane Dirac–Born–Infeld action, up to a Chern–Simons term,

$N_c$1

with

$N_c$2

Expanding to quadratic order yields the effective five-dimensional gauge action

$N_c$3

where

$N_c$4

This is the basic mesonic sector used to extract normalizable vector and axial-vector modes and their couplings to glueballs (Parganlija, 2015).

The pseudoscalar sector is encoded in the holonomy of the holographic gauge field,

$N_c$5

or equivalently, in another standard convention,

$N_c$6

Vector and axial-vector mesons arise from higher normalizable modes of $N_c$7. This provides the usual WSS interpretation of pions, $N_c$8, $N_c$9, axial-vector mesons, and their tower excitations as Kaluza–Klein modes of the flavor-brane gauge field (Brünner et al., 2015).

The Chern–Simons sector is essential for anomaly physics. It generates the Wess–Zumino–Witten structure, vector-meson dominance, and the Witten–Veneziano mechanism. The anomalous singlet mass is

$\lambda$0

For $\lambda$1, $\lambda$2 MeV, and $\lambda$3, one obtains

$\lambda$4

together with

$\lambda$5

and

$\lambda$6

This sector is one of the clearest examples of the model reproducing a quantitatively nontrivial QCD feature from a higher-dimensional origin (Brünner et al., 2015).

Meson phenomenology is semi-quantitatively successful in a number of channels. For example,

$\lambda$7

were quoted as being in good agreement with experiment. Radiative decays are also described through holographic vector-meson dominance rather than by a separate phenomenological ansatz (Brünner et al., 2015).

3. Glueballs and the bulk origin of hadronic singlets

In the WSS model, glueballs are normalizable bulk fluctuations rather than added four-dimensional fields. Scalar and tensor glueballs arise from dilaton and graviton modes in the confining D4 background, while the pseudoscalar glueball is associated with fluctuations of the Ramond–Ramond one-form $\lambda$8, tying it directly to the holographic treatment of the $\lambda$9 anomaly (Brünner et al., 2017).

A major structural feature of the scalar sector is the distinction between the lowest “exotic” scalar mode and the next, predominantly dilatonic scalar mode. The masses quoted in the WSS spectrum are

$M_{\rm KK}$0

with the tensor $M_{\rm KK}$1 degenerate with the dilatonic scalar at the supergravity level (Brünner et al., 2015). The exotic scalar involves an unusual graviton polarization associated with the compactification direction; several WSS phenomenology papers therefore argue that it should be discarded as the physical QCD scalar glueball candidate, while the dilatonic scalar is taken as the relevant lightest scalar glueball mode (Brünner et al., 2015).

Glueball–meson couplings descend from the same D8-brane action that governs the meson sector. For the tensor glueball, the low-energy pion coupling is

$M_{\rm KK}$2

For the dilatonic scalar,

$M_{\rm KK}$3

showing the characteristic large-$M_{\rm KK}$4, large-$M_{\rm KK}$5 suppression of glueball widths (Rebhan, 2016).

The pseudoscalar glueball is exceptional. In WSS there is no direct coupling of the Ramond–Ramond one-form fluctuation to the flavor D8-branes, so the leading interaction is a $M_{\rm KK}$6-$M_{\rm KK}$7-$M_{\rm KK}$8 vertex involving a scalar glueball, a pseudoscalar glueball, and the singlet pseudoscalar. This implies that the pseudoscalar glueball is expected to be a very narrow state, with phenomenology dominated by

$M_{\rm KK}$9

rather than by direct leading-order decays into ordinary meson pairs (Brünner et al., 2017).

Radiative decays extend the same logic. The photon enters the WSS model through asymptotic flavor sources and vector-meson dominance. In this framework, scalar, tensor, and pseudoscalar glueballs all acquire calculable two-photon couplings. The resulting two-photon widths are not tiny: the 2023 analysis obtained keV-scale $\lambda$0 widths for scalar, tensor, and pseudoscalar glueballs, and argued that the observed two-photon rate of $\lambda$1 is not too large to permit a glueball interpretation within WSS (Hechenberger et al., 2023).

4. Scalar, tensor, and pseudoscalar glueball phenomenology

The scalar-glueball program in WSS has focused on the comparison between $\lambda$2 and $\lambda$3. In the original chiral-limit analysis, the dilatonic scalar already gave a better account of $\lambda$4 than of $\lambda$5, whereas the exotic scalar was too light and too broad (Brünner et al., 2015). Finite quark masses sharpen this distinction.

The finite-mass deformation introduces a nonlocal mass term in the holographic direction and motivates an effective glueball coupling to pseudoscalar mass terms. The key result is not “chiral suppression” of $\lambda$6, but what the authors call “nonchiral enhancement”: the chiral-limit coupling to Nambu–Goldstone bosons is already nonzero, and finite quark masses add an extra contribution that enhances decays into heavier pseudoscalars. The central multiplicative factor for $\lambda$7 is

$\lambda$8

With this mechanism, the WSS prediction

$\lambda$9

was compared with the experimental value

$N_c$0

while

$N_c$1

was compared with

$N_c$2

Within this framework, $N_c$3 is favored as the predominantly gluonic scalar state, whereas $N_c$4 is disfavored as a nearly unmixed glueball (Brünner et al., 2015).

The tensor glueball behaves differently. At the uncorrected holographic mass,

$N_c$5

the chiral-limit partial width

$N_c$6

makes it appear rather narrow. However, this is below the lattice-QCD window

$N_c$7

When the tensor mass is extrapolated upward, the vector-vector channels

$N_c$8

open and dominate. At

$N_c$9

the quoted decay widths were

$M_{\rm KK}$0

$M_{\rm KK}$1

with total width

$M_{\rm KK}$2

The model therefore predicts that a nearly pure tensor glueball in the lattice-favored mass region is extremely broad and difficult to isolate experimentally (Parganlija, 2015).

A recurrent qualitative picture emerges across the WSS glueball literature. The predominantly dilatonic scalar can be compatible with $M_{\rm KK}$3 once finite pseudoscalar masses are included; the tensor glueball is likely too broad at realistic masses; and the pseudoscalar glueball remains narrow because its leading couplings are structurally restricted by the Ramond–Ramond sector (Rebhan, 2016).

5. Baryons, generalized Skyrmions, and dense matter

Baryons in WSS are 5D instantons of the flavor gauge theory; equivalently, they are wrapped D4-brane baryon vertices, and at low energy they project to four-dimensional Skyrmions. This is one of the most distinctive aspects of the model, because the baryonic sector is not introduced separately but emerges from the same flavor-brane action that yields mesons (Bartolini et al., 2017).

A refinement of the usual pion-only truncation shows that integrating out the Abelian vector mode associated mainly with the $M_{\rm KK}$4 meson and its tower generates a sextic term,

$M_{\rm KK}$5

where

$M_{\rm KK}$6

is the topological current. The resulting low-energy action is a generalized Skyrme model,

$M_{\rm KK}$7

In the small-$M_{\rm KK}$8 limit the sextic term becomes dominant; with a pion mass term included, the model reduces to the BPS Skyrme model in that limit (Bartolini et al., 2017).

At finite baryon density, many WSS studies use a homogeneous smeared-instanton ansatz. In this approximation the gauge profile must be discontinuous at the infrared tip in order to support nonzero baryon density. This gives rise to a subtlety in the Chern–Simons boundary term. A 2023 analysis showed that thermodynamics can be made consistent with or without the extra infrared boundary contribution, but argued that the preferred prescription is to discard it, because only then does the baryon density extracted from the UV current agree with the topological instanton density carried by the ansatz (Bartolini et al., 2023).

The model has also been pushed into quantitative neutron-star phenomenology. Using the homogeneous baryonic ansatz, a recalibration was proposed that fixes $M_{\rm KK}$9 and $\lambda=g_{\mathrm{YM}}^2N_c$0 not from meson observables but from the physical saturation density and symmetry energy. The two benchmark fits quoted were

$\lambda=g_{\mathrm{YM}}^2N_c$1

and

$\lambda=g_{\mathrm{YM}}^2N_c$2

After hybridization with low-density nuclear equations of state, the resulting neutron-star models yielded

$\lambda=g_{\mathrm{YM}}^2N_c$3

compatible with PSR J0952-0607 and with radius and tidal-deformability constraints. The same work emphasized, however, that the standard meson-sector fit performs badly in the baryonic sector and that the homogeneous ansatz is unreliable near saturation density (Bartolini et al., 2023).

The D0-D4 deformation provides another baryonic extension. In the three-flavor D0-D4/WSS setup, the smeared D0-brane density is dual to a nonzero topological background

$\lambda=g_{\mathrm{YM}}^2N_c$4

parameterized by $\lambda=g_{\mathrm{YM}}^2N_c$5. Quantization of the baryon collective coordinates in this background led to the conclusion that stable baryons require

$\lambda=g_{\mathrm{YM}}^2N_c$6

and that increasing D0 charge reduces baryon splittings, suggesting that sufficiently large $\lambda=g_{\mathrm{YM}}^2N_c$7-like deformation or glue condensate destabilizes baryons (Cai et al., 2017).

6. Thermal phases, chemical potentials, external fields, and model status

At finite temperature the WSS model distinguishes confinement/deconfinement from chiral symmetry breaking/restoration by the topology of the Euclidean geometry and by the D8-brane embedding. In the low-temperature background only connected embeddings exist, so chiral symmetry is broken. In the high-temperature geometry, connected and horizon-reaching embeddings compete, allowing chiral restoration. With imaginary quark-number chemical potential, Roberge–Weiss transitions occur only in the high-temperature, chiral-symmetry-restored phase; the low-temperature chirally broken phase is smooth in imaginary chemical potential. The chiral transition was found to be first order, and where it meets Roberge–Weiss lines the intersections are triple points (Rafferty, 2011).

An axial chemical potential in the low-temperature phase drives another instability. In the Maxwell–Chern–Simons truncation, unstable zero modes appear above

$\lambda=g_{\mathrm{YM}}^2N_c$8

while in the DBI–Chern–Simons treatment the critical density becomes

$\lambda=g_{\mathrm{YM}}^2N_c$9

Above this threshold, the confining chirally broken vacuum is destabilized toward a spatially modulated phase with a helical vector current condensate. The Chern–Simons term drives the instability; the DBI nonlinearity raises the threshold but does not remove the phase for the physical WSS coupling (Bayona et al., 2011).

External magnetic fields promote chiral symmetry breaking. In the deconfined phase, the magnetic field bends the flavor branes more strongly toward joining, raising the chiral-restoration temperature above its zero-field value, with saturation at large field strength. This was identified with magnetic catalysis in the holographic WSS context (0803.0038).

The D0-D4 deformation further extends the model to $N_f\ll N_c$0-dependent backgrounds. In the Schwinger-effect analysis, the imaginary part of the probe D8-brane Euler–Heisenberg action decreases as the D0-brane density increases, and the critical electric field rises with the parameter $N_f\ll N_c$1. This was interpreted as suppression of electromagnetic vacuum decay by a topological/gluonic background that makes the vacuum heavier (Cai et al., 2016).

As a description of dense matter with baryon and isospin chemical potentials, WSS admits multiple connected and disconnected D8-brane phases. In the two-flavor analysis with baryon number and $N_f\ll N_c$2 isospin, connected phases represent chirally broken matter, while parallel disconnected embeddings correspond to chiral restoration. The competition between dense connected and disconnected deconfined phases was found to exhibit temperature-driven transitions, including re-entrant behavior in some parameter ranges (Sethi, 2019).

These developments underscore both the scope and the limits of the model. The WSS framework has been used to compute meson observables, glueball decay patterns, baryon spectra, dense-matter equations of state, Roberge–Weiss transitions, axial-density instabilities, magnetic catalysis, and radiative glueball widths from a common top-down construction. At the same time, its quantitative comparison to QCD depends on large-$N_f\ll N_c$3 and large-$N_f\ll N_c$4 extrapolation, probe-flavor dynamics, Kaluza–Klein contamination, assumptions about finite quark masses, and, in several sectors, phenomenological mass shifts or sector-specific recalibrations (Rebhan, 2014).

Within those bounds, the model’s most robust qualitative outputs are the geometric realization of chiral symmetry breaking, the common higher-dimensional origin of mesons and glueballs, the large-$N_f\ll N_c$5 scaling of hadronic widths, the anomaly structure encoded by the Chern–Simons and Ramond–Ramond sectors, and the emergence of baryons as instantons of the flavor gauge theory. A plausible implication is that the WSS model is best regarded not as an exact dual of QCD but as a highly constrained holographic laboratory in which infrared hadron dynamics, anomaly physics, and dense-matter phenomena can be computed in a unified framework.