Soft-wall AdS/QCD Model Overview

• Soft-wall AdS/QCD is a holographic framework that uses a smooth dilaton profile to regulate infrared dynamics in a five-dimensional setting.
• It replicates key QCD features such as linear Regge trajectories, chiral symmetry breaking, and a massless pion in the chiral limit.
• The model blends large-Nc QCD sum-rule techniques with thermal and dense matter extensions to achieve a unified description.

Soft-wall AdS/QCD is a bottom-up holographic framework in which the infrared sector of a five-dimensional asymptotically AdS description is regulated by a smooth dilaton background rather than a hard cutoff, with the principal aim of reproducing confinement-related features such as linear Regge trajectories while incorporating chiral symmetry breaking and other low-energy properties of QCD (Sachan et al., 2011, Kapusta et al., 2010). In standard formulations the geometry extends to $z\to\infty$, but a factor such as $e^{-\Phi(z)}$ suppresses the large-$z$ region smoothly; in light-front-holographic formulations, a positive-sign dilaton metric yields an effective harmonic confining potential, a massless pion in the chiral limit, and linear trajectories in both radial and orbital quantum numbers (Brodsky et al., 2010, Brodsky et al., 2010). Subsequent work recast the model as a five-dimensional rewriting of planar QCD sum-rule phenomenology, reconstructed scalar potentials supporting soft-wall backgrounds, constrained the infrared behavior by simultaneous meson and nucleon spectroscopy, and extended the framework to thermodynamics, dense matter, and chiral phase transitions (Afonin, 2010, He et al., 2011, Fang et al., 2019).

1. Conceptual foundations

The starting point is usually AdS$_5$ with metric

$$ds^2=\frac{R^2}{z^2}\left(\eta_{\mu\nu}dx^\mu dx^\nu-dz^2\right),$$

where the fifth-dimensional coordinate $z$ plays the role of a holographic scale: small $z$ corresponds to ultraviolet physics and large $z$ to infrared physics (Brodsky et al., 2010, Brodsky et al., 2010). In the soft-wall model, conformal invariance is broken smoothly by a dilaton profile rather than by a hard infrared cutoff. In one common convention, the action carries $e^{-\Phi(z)}$ with $\Phi(z)=c z^2$, so the large-$e^{-\Phi(z)}$0 region is exponentially suppressed while the spacetime itself still extends to $e^{-\Phi(z)}$1 (Sachan et al., 2011). In another convention, especially in light-front holography, the soft wall is described through a positive-sign dilaton metric $e^{-\Phi(z)}$2, which produces a confining potential with a minimum at $e^{-\Phi(z)}$3 and rising behavior at large $e^{-\Phi(z)}$4 (Brodsky et al., 2010, Brodsky et al., 2010).

A central result of the light-front-holographic formulation is the identification $e^{-\Phi(z)}$5, where

$e^{-\Phi(z)}$6

is the invariant light-front transverse separation of constituents (Brodsky et al., 2010). The corresponding light-front Schrödinger equation,

$e^{-\Phi(z)}$7

with

$e^{-\Phi(z)}$8

gives the meson spectrum

$e^{-\Phi(z)}$9

so that $z$0 for $z$1 and also for $z$2 (Brodsky et al., 2010). For $z$3, one obtains $z$4 in the massless-quark limit (Brodsky et al., 2010).

A complementary interpretation treats bottom-up holographic models as a five-dimensional rewriting of large-$z$5 QCD, where an infinite tower of narrow mesons in each channel is packaged into a single 5D field with a Sturm–Liouville spectrum (Afonin, 2010). Under mild assumptions, that analysis classifies the 5D backgrounds leading to simple Regge spectra and argues that the most phenomenologically consistent model is the soft wall, with a preference for the positive-sign dilaton background (Afonin, 2010). This suggests that the soft-wall model is not only an AdS/CFT-inspired construction but also an alternative language for planar QCD sum-rule phenomenology.

2. Bulk actions, background fields, and representative constructions

A standard meson-sector action is written in string frame as

$z$6

with bifundamental scalar $z$7, left and right gauge fields $z$8, and vacuum expectation value

$z$9

The bulk gauge fields encode vector and axial-vector modes through

$_5$0

(Bartz et al., 2013). In improved meson models the scalar interaction is often quartic, and the same general structure is written as

$_5$1

or with $_5$2-dependent scalar mass and quartic coupling in pure AdS$_5$3 (Kelley et al., 2010, Cui et al., 2013).

For the background sector, one frequently passes to Einstein frame. In one two-field form,

$_5$4

and in the three-field extension the glueball condensate $_5$5 is added as an additional scalar (Bartz et al., 2013). A systematic potential-reconstruction program shows how to build $_5$6 for prescribed soft-wall and chiral backgrounds, starting from the Einstein-frame action

$_5$7

and an exact AdS$_5$8 string-frame metric (Kapusta et al., 2010).

The main bottom-up variants represented in the literature can be organized succinctly as follows.

Construction	Distinguishing ingredient	Representative papers
Standard soft wall	Quadratic dilaton and smooth IR suppression	(Sachan et al., 2011, Kapusta et al., 2010)
Light-front-holographic soft wall	Positive-sign dilaton metric and $_5$9 map	(Brodsky et al., 2010, Brodsky et al., 2010)
Constrained IR background	$$ds^2=\frac{R^2}{z^2}\left(\eta_{\mu\nu}dx^\mu dx^\nu-dz^2\right),$$0, $$ds^2=\frac{R^2}{z^2}\left(\eta_{\mu\nu}dx^\mu dx^\nu-dz^2\right),$$1, $$ds^2=\frac{R^2}{z^2}\left(\eta_{\mu\nu}dx^\mu dx^\nu-dz^2\right),$$2 from mesons and nucleons	(He et al., 2011, Liu et al., 2012)
Three-field background	Dilaton, chiral condensate, and glueball condensate $$ds^2=\frac{R^2}{z^2}\left(\eta_{\mu\nu}dx^\mu dx^\nu-dz^2\right),$$3	(Bartz et al., 2013, Bartz et al., 2012)
Improved thermal background	Einstein–dilaton geometry fitted to lattice thermodynamics	(Fang et al., 2019)

These constructions share the same broad objective—smooth infrared breaking of conformal symmetry—but differ in how much of the background is imposed, reconstructed, or solved dynamically. A plausible implication is that “soft-wall AdS/QCD model” denotes a family of closely related bottom-up frameworks rather than a single universal action.

3. Chiral symmetry breaking and infrared engineering

A recurring difficulty of the original soft-wall model is that the combination of a purely quadratic scalar potential and the standard positive quadratic dilaton does not generate spontaneous chiral symmetry breaking. In the two-flavor analysis,

$$ds^2=\frac{R^2}{z^2}\left(\eta_{\mu\nu}dx^\mu dx^\nu-dz^2\right),$$4

at zero temperature remains linear, and the exact $$ds^2=\frac{R^2}{z^2}\left(\eta_{\mu\nu}dx^\mu dx^\nu-dz^2\right),$$5 solution in the positive quadratic dilaton background gives

$$ds^2=\frac{R^2}{z^2}\left(\eta_{\mu\nu}dx^\mu dx^\nu-dz^2\right),$$6

so $$ds^2=\frac{R^2}{z^2}\left(\eta_{\mu\nu}dx^\mu dx^\nu-dz^2\right),$$7 and $$ds^2=\frac{R^2}{z^2}\left(\eta_{\mu\nu}dx^\mu dx^\nu-dz^2\right),$$8 in the chiral limit (Chelabi et al., 2015). That work concludes that realistic chiral dynamics requires both a nonlinear scalar potential,

$$ds^2=\frac{R^2}{z^2}\left(\eta_{\mu\nu}dx^\mu dx^\nu-dz^2\right),$$9

and a dilaton profile that is negative in the UV/intermediate region but positive quadratic in the IR,

$z$0

(Chelabi et al., 2015). In the two-flavor case this gives a second-order transition in the chiral limit and a crossover for any finite quark mass; in the three-flavor case the cubic term induced by the t’Hooft determinant interaction drives first-order behavior at small masses (Chelabi et al., 2015, Fang et al., 2018).

A different line of improvement keeps the geometry as pure $z$1 but moves the necessary infrared deformation into the matter sector. The infrared-improved soft-wall model for mesons uses

$z$2

a $z$3-dependent quartic coupling

$z$4

and a $z$5-dependent scalar mass $z$6 whose UV limit approaches $z$7 and whose IR asymptotics are fixed by the scalar equation of motion (Cui et al., 2013). The scalar VEV is parameterized as

$z$8

with UV source/VEV behavior and linear IR growth (Cui et al., 2013). This arrangement is designed to realize linear confinement and chiral symmetry breaking within one simple framework.

A further constraint arises when mesons and nucleons are fitted simultaneously. Requiring linear Regge behavior in both sectors and equal asymptotic slopes for vector and axial-vector mesons leads to the infrared conditions

$z$9

while the UV remains

$z$0

(He et al., 2011). These asymptotics motivate the rational interpolating forms

$z$1

used later in the quartic scalar model of mesons and nucleons (He et al., 2011, Liu et al., 2012).

The most explicit structural critique of two-field soft-wall backgrounds appears in the three-field construction. In the two-field case one has the potential-independent relation

$z$2

and imposing the soft-wall infrared forms

$z$3

forces

$z$4

which predicts a vector/axial-vector mass splitting about $z$5 times too large (Bartz et al., 2013). Introducing a glueball condensate $z$6, dual to the gluonic operator $z$7, changes the key background equation to

$z$8

so the chiral field no longer bears the full burden of sourcing the quadratic dilaton (Bartz et al., 2013). The UV and IR asymptotics are then

$z$9

and

$z$0

(Bartz et al., 2013). The authors explicitly state that constructing a globally valid scalar potential consistent with both UV and IR limits remains open, but the three-field parametrization resolves the major phenomenological inconsistency of the two-field setup (Bartz et al., 2013).

4. Spectroscopy of mesons and baryons

The original phenomenological appeal of the soft wall lies in spectroscopy. In the light-front-holographic version, the meson mass formula

$z$1

gives linear trajectories with equal slopes in the radial quantum number $z$2 and orbital angular momentum $z$3, and the baryon sector satisfies

$z$4

(Brodsky et al., 2010). The same framework yields explicit proton wavefunctions,

$z$5

and predicts a massless pion for zero quark mass (Brodsky et al., 2010, Brodsky et al., 2010).

In meson-sector bottom-up models, the fluctuation equations are typically cast into Sturm–Liouville or Schrödinger-like form. The vector sector depends on the dilaton and warp factor, while the axial sector contains an additional effective mass term proportional to $z$6, which is why the infrared behavior of the chiral background is crucial (Bartz et al., 2013). The constrained-IR model with quartic scalar interaction reports good agreement with experimental meson and nucleon spectra once the background functions are chosen as

$z$7

and a running bulk spinor mass is introduced for the nucleons (Liu et al., 2012).

The pseudoscalar sector required additional clarification. In the modified soft-wall model with quartic scalar interaction and a nontrivial scalar vacuum profile $z$8, the pion can be represented either exponentially,

$z$9

or linearly,

$e^{-\Phi(z)}$0

The two representations are related by

$e^{-\Phi(z)}$1

and become exactly equivalent once the scalar background equation is used (Kelley et al., 2010). Solving the full coupled system, rather than reducing it prematurely to a single equation for $e^{-\Phi(z)}$2, produces a realistic pion tower with a very light ground state, a large gap to the first radial excitation, and a roughly linear radial trajectory (Kelley et al., 2010). The Gell-Mann–Oakes–Renner relation,

$e^{-\Phi(z)}$3

also follows naturally (Kelley et al., 2010).

Infrared-improved meson models sharpen these results. One such model reports scalar, pseudoscalar, vector, and axial-vector masses

$e^{-\Phi(z)}$4

for the scalar tower,

$e^{-\Phi(z)}$5

for vectors, and

$e^{-\Phi(z)}$6

for axial-vectors, with the lightest scalar around $e^{-\Phi(z)}$7 MeV and a reasonable space-like pion form factor (Cui et al., 2013). In the three-field soft-wall background, a least-squares fit to the $e^{-\Phi(z)}$8 and $e^{-\Phi(z)}$9 spectra gives

$\Phi(z)=c z^2$0

and the quartic coupling

$\Phi(z)=c z^2$1

which removes the unbounded-from-below scalar potential of earlier work and avoids a virtual $\Phi(z)=c z^2$2 ground state (Bartz et al., 2013).

Not all observables are reproduced equally well in the standard prescription. For vector-meson decay constants, the original soft wall predicts

$\Phi(z)=c z^2$3

and a radial-independent decay constant

$\Phi(z)=c z^2$4

which is inconsistent with heavy-quarkonium data, where the decay constants decrease with excitation number (Braga et al., 2015). A modified prescription evaluates the current-current correlator at a finite radial position $\Phi(z)=c z^2$5, with bulk-to-boundary propagator

$\Phi(z)=c z^2$6

thereby introducing a second ultraviolet scale $\Phi(z)=c z^2$7 and yielding decreasing decay constants for heavy vector mesons (Braga et al., 2015).

5. Thermal, dense, and phase-structure extensions

Finite-temperature and finite-density soft-wall models typically introduce a black-hole background and a bulk $\Phi(z)=c z^2$8 field dual to quark number. In one standard treatment, the gravity action is

$\Phi(z)=c z^2$9

with

$e^{-\Phi(z)}$00

thermal AdS as the confined phase,

$e^{-\Phi(z)}$01

and AdS black hole as the deconfined phase,

$e^{-\Phi(z)}$02

(Sachan et al., 2011). At $e^{-\Phi(z)}$03, the Hawking–Page-type transition occurs at

$e^{-\Phi(z)}$04

and the quark number susceptibility is

$e^{-\Phi(z)}$05

which reduces at high temperature to

$e^{-\Phi(z)}$06

(Sachan et al., 2011). That paper explicitly notes that the results are valid only for small chemical potential because backreaction is neglected (Sachan et al., 2011).

The phase boundary can also be modeled spectrally rather than thermodynamically. In an extended soft-wall model, vector mesons are identified with normalizable bulk modes, and their disappearance at $e^{-\Phi(z)}$07 is adjusted to track the freeze-out or crossover curve at small $e^{-\Phi(z)}$08 (Zöllner et al., 2016). The background metric is

$e^{-\Phi(z)}$09

with generalized dilaton

$e^{-\Phi(z)}$10

and a phenomenological blackness function engineered to reproduce thermal-gas versus black-hole competition (Zöllner et al., 2016). A plausible implication is that in soft-wall practice the deconfinement line can be defined either from free-energy competition or from the disappearance of normalizable hadronic modes, depending on the construction.

Chiral transitions require a separate scalar-sector analysis. In the improved soft-wall model with nonlinear scalar potential and interpolating dilaton, the two-flavor case exhibits a second-order transition in the chiral limit and a crossover for finite quark mass, while the three-flavor case exhibits first-order behavior in the chiral limit and at small finite masses, with a critical mass separating first-order and crossover regimes (Chelabi et al., 2015). The $e^{-\Phi(z)}$11-flavor extension uses

$e^{-\Phi(z)}$12

plus a determinant term $e^{-\Phi(z)}$13, and yields a Columbia-plot structure with a tricritical point on the $e^{-\Phi(z)}$14 boundary at

$e^{-\Phi(z)}$15

(Fang et al., 2018).

A more realistic thermal background is obtained by solving an Einstein–dilaton system with nontrivial dilaton potential, rather than imposing AdS-Schwarzschild. In that model the string-frame metric is

$e^{-\Phi(z)}$16

the Einstein-frame action is

$e^{-\Phi(z)}$17

and the potential

$e^{-\Phi(z)}$18

is tuned to reproduce the two-flavor lattice equation of state (Fang et al., 2019). On top of that background, an improved soft-wall flavor action with dilaton–scalar coupling

$e^{-\Phi(z)}$19

produces a crossover chiral transition, although the chiral transition temperature remains too high relative to lattice (Fang et al., 2019).

Finite isospin density exposes additional limitations. In a soft-wall pion-condensation study with improved dilaton

$e^{-\Phi(z)}$20

and AdS-Reissner–Nordström-like background,

$e^{-\Phi(z)}$21

the phase transition into and out of the pion-condensed phase is found to be first order both at small and large $e^{-\Phi(z)}$22 (Lv et al., 2018). The same paper states that the disagreement with lattice behavior at small $e^{-\Phi(z)}$23 indicates that a full back-reaction model including the interaction of gluodynamics and chiral dynamics is likely necessary (Lv et al., 2018).

6. Conventions, controversies, and open problems

A persistent source of confusion in the literature is the sign convention for the dilaton. In the thermal soft-wall model and many meson-sector constructions, the action carries $e^{-\Phi(z)}$24 with $e^{-\Phi(z)}$25 (Sachan et al., 2011). In the light-front-holographic formulation, by contrast, the defining statement is that the confining choice is a positive-sign dilaton metric,

$e^{-\Phi(z)}$26

opposite in sign to the original Karch–Katz–Son–Stephanov convention (Brodsky et al., 2010, Brodsky et al., 2010). Afonin’s classification of five-dimensional large-$e^{-\Phi(z)}$27 rewritings also argues for a preference for the positive-sign dilaton background (Afonin, 2010). These are not merely superficial notational differences: they arise in different frames and different effective formulations, and each comes with its own interpretation of the confining background.

Another recurring issue is the bottom-up status of the framework. Several papers emphasize that soft-wall AdS/QCD is phenomenological rather than derived from a full top-down string construction (Sachan et al., 2011, Chelabi et al., 2015, Fang et al., 2019). The large-$e^{-\Phi(z)}$28 rewriting program goes further and argues that many features of AdS/QCD can be understood without invoking the full AdS/CFT correspondence at all, as a 5D reformulation of planar QCD sum rules (Afonin, 2010). This suggests that the success of the soft wall may depend more on its spectral and symmetry structure than on a strict microscopic duality.

Dynamic AdS/QCD sharpens the question of what a soft wall means physically. In that framework the running constituent mass $e^{-\Phi(z)}$29 itself generates an infrared wall through

$e^{-\Phi(z)}$30

so if $e^{-\Phi(z)}$31, meson dynamics never reaches $e^{-\Phi(z)}$32 (Evans et al., 2015). The natural outcome is then square-well-like behavior,

$e^{-\Phi(z)}$33

rather than Regge behavior. To obtain

$e^{-\Phi(z)}$34

one must force $e^{-\Phi(z)}$35 to fall sharply in the IR, for example by taking

$e^{-\Phi(z)}$36

or by other engineered profiles (Evans et al., 2015). That paper concludes that Regge-producing soft-wall behavior requires meson physics to remain sensitive to scales below the quark on-shell mass, and therefore that the meson spectrum is highly sensitive to IR decoupling (Evans et al., 2015). A plausible implication is that “soft-wall behavior” and “dynamical constituent-mass screening” are not automatically compatible.

Several open problems remain explicit in the literature. The three-field model resolves the structural two-field obstruction but does not yet supply a globally valid scalar potential consistent with both UV and IR limits (Bartz et al., 2013). The pion-condensation study indicates that full backreaction may be required to capture low-$e^{-\Phi(z)}$37 physics correctly (Lv et al., 2018). The Einstein–dilaton thermal model shows that deconfinement-related thermodynamics and chiral restoration can be treated in one framework, but still in the probe approximation for the flavor sector (Fang et al., 2019). Across the literature, the soft-wall AdS/QCD model therefore appears less as a finished theory than as an evolving family of constrained effective constructions whose defining achievement is the simultaneous modeling of smooth infrared confinement, Regge behavior, and chiral dynamics within a five-dimensional holographic language (Kapusta et al., 2010, Bartz et al., 2013).