Scale-Separated AdS Solutions

• Scale-separated AdS solutions are vacua where the lightest Kaluza–Klein modes far exceed the AdS curvature scale, enabling effective lower-dimensional dynamics.
• Classical no-go theorems typically enforce m_KK ~ m_AdS unless mechanisms like negative-tension sources or higher-order corrections are introduced.
• Explicit constructions such as DGKT vacua use unbounded flux scaling and orientifold planes to achieve a parametric hierarchy while maintaining control over moduli and corrections.

A scale-separated AdS solution is a supergravity (or full string-theoretic) vacuum with anti-de Sitter (AdS) spacetime in which the mass scale of the lightest Kaluza–Klein (KK) excitations of the compact internal space is parametrically larger than the AdS curvature scale, i.e., where $m_{\text{KK}}^2 \gg |\Lambda|$ or equivalently $L_{\text{AdS}} / L_{\text{KK}} \gg 1$ in appropriate limits. Such vacua provide genuine lower-dimensional effective theories with AdS geometry well below the cutoff set by the lightest KK modes. The problem of realizing scale-separated AdS solutions in string theory or supergravity is central to the viability of effective AdS holography, landscape studies, and Swampland physics.

1. Definition and Necessary Conditions for Scale Separation

A solution to $D$-dimensional supergravity (often with $D=10,11$) is said to exhibit scale separation if, after compactification on a compact internal manifold $M_{d}$ of dimension $d = D - d_{\text{ext}}$ (where $d_{\text{ext}}$ is the dimension of the AdS spacetime), the ratio

$$\Delta \equiv \frac{m_{\rm KK}}{m_{\rm AdS}} = m_{\rm KK} L_{\rm AdS} \gg 1$$

in some parametric regime (typically a large-flux limit or a scaling of geometric moduli). Here $m_{\rm KK}$ is the lightest nontrivial Laplace eigenvalue on $M_{d}$ and $m_{\rm AdS}^2 = |\Lambda|$ is the AdS curvature. True scale separation requires that this ratio can be made arbitrarily large while maintaining control over string coupling and internal volumes. The construction must also ensure all other moduli are stabilized and that higher-derivative or quantum corrections remain suppressed.

The primary necessary conditions for scale separation are:

• A mechanism to generate a small cosmological constant $|\Lambda|$ compared to the KK scale, without losing control over the low-energy effective description.
• Stabilization of all geometric and string-theoretic moduli via fluxes, curvature, orientifold planes, or non-perturbative effects.

2. Classical No-go Theorems and Curvature Obstructions

A series of classical no-go theorems, culminating in extensions of the Maldacena–Núñez argument, imply that in pure $D$-dimensional gravity with smooth compact internal space and only fluxes,

$|R_{d}| / |R_{d_{\rm ext}}| = \mathcal{O}(1)$

where $R_{d}$ is the internal Ricci scalar and $R_{d_{\rm ext}}$ is the AdS curvature. As a consequence, the lightest KK mass and the AdS radius are generically of the same order: $m_{\text{KK}}^2 \sim |\Lambda|$ with no parametric separation unless special ingredients are introduced (Gautason et al., 2015, Font et al., 2019, Tsimpis, 2012, Lust et al., 2020). The only exceptions arise when:

• Negative-tension sources (e.g., orientifold planes) enter at leading order in the potential;
• Strong dilaton gradients are allowed (often outside supergravity control);
• Higher-derivative or quantum effects become important.

Thus, compactifications on internal spaces with leading isotropic curvature (e.g., Freund–Rubin type $AdS_{d_{\rm ext}} \times M_{d}$ with $M_{d}$ an Einstein manifold) always yield $m_{\rm KK} \sim m_{\rm AdS}$ and preclude true scale separation (Gautason et al., 2015, Tsimpis, 2012).

3. Orientifold and Flux Mechanisms for Achieving Scale Separation

Parametric scale separation can be achieved classically within controllable corners of string theory when:

• Orientifold planes (O$p$ with $p<7$) provide a leading negative contribution to the scalar potential, with two types of fluxes (NSNS and RR) cancelling their tadpoles (Tringas et al., 21 Apr 2025);
• An additional “unbounded” RR-flux can be scaled to large values, pushing the internal volume $\mathcal{V}$ to infinity and string coupling $g_s \to 0$, resulting in the hierarchy

$$m_{\rm KK}/m_{\rm AdS} \sim g_s^{1/2} \mathcal{V}^{-(13-2p)/12} \rightarrow \infty$$

for $p<7$. Representative examples include:

• Massive IIA on CY or toroidal orientifolds with O6 planes and large $F_4$ flux [DeWolfe–Giddings–Kachru–Taylor, (Shiu et al., 2022, Apers et al., 2022, Carrasco et al., 2023)];
• Type IIB with intersecting O5/O7 planes on SU(2)-structure or solmanifolds (Petrini et al., 2013, Hemelryck, 7 Feb 2025);
• Massive IIA or IIB on $G_2$-structure sevenfolds with O2, O6, O5, or dual NS5 planes, often on nilmanifolds or solvmanifolds (Farakos et al., 2020, Miao et al., 16 Sep 2025, Tringas et al., 11 Nov 2025).

The scaling limit is realized by sending an unconstrained flux $N \rightarrow \infty$, with all tadpole-cancelling fluxes kept fixed. The parametric control is then achieved by scaling

$$g_s \sim N^{-c_1},\quad \mathcal{V} \sim N^{c_2},\quad L_{\rm AdS} \sim N^{c_3},\quad m_{\rm KK} \sim N^{-c_4}$$

with exponents determined by the details of the flux background.

4. Explicit Constructions and T-dual Families

Within this framework, several families of scale-separated AdS vacua have been constructed:

• DGKT Vacua and Generalizations: In the massive IIA “DGKT” solutions, the AdS scale is set by the RR-flux $N$,

$$L_{\rm AdS} \sim N^{3/4},\quad L_{\rm KK} \sim N^{1/4},\quad L_{\rm AdS}/L_{\rm KK} \sim N^{1/2} \rightarrow \infty$$

with all moduli stabilized, $g_s \to 0$, and large internal volume (Shiu et al., 2022, Carrasco et al., 2023, Andriot et al., 2023). Extensions to anisotropic orbifolds allow for even more general flux scalings (Tringas, 2023).

• AdS$_3$ Vacua on G$_2$-Structure: Compactifications of massive IIA or dual IIB/type I on seven-manifolds with $G_2$- or co-calibrated $G_2$-structure, with the crucial ingredient that certain “bulk” (cohomologically nontrivial) three-form fluxes are unbounded and not fixed by tadpoles. This yields solutions with

$$L_{\rm AdS} \sim N^{a},\quad L_{\rm KK} \sim N^{b},\quad L_{\rm AdS}/L_{\rm KK} \sim N^{a-b} \gg 1$$

for appropriate choices (Farakos et al., 2020, Miao et al., 16 Sep 2025, Hemelryck, 7 Feb 2025, Tringas et al., 11 Nov 2025). S-duality relates some type I solutions to classical heterotic SO(32) scale-separated vacua with only NSNS sector and gravitational instantons (Tringas et al., 11 Nov 2025).

• No-scale and Non-geometric Examples: In simple no-scale limits (e.g., absence of Romans mass in IIA on G$_2$-orbifolds), the vacua can become more sophisticated or non-geometric (e.g., IIB backgrounds with non-geometric Q-flux). However, full scale separation still correlates with unrestricted flux parameters (Carrasco et al., 2023, Miao et al., 16 Sep 2025).
• IIB SU(2)-structure O5/O7 Vacua: Type IIB with intersecting O5 and O7 planes and SU(2)-structure internal manifolds allow for exact algebraic solutions with scale separation achieved by scaling certain flux/torsion singlets (Petrini et al., 2013).

A key universal feature is that exact Kodaira–Spencer deformations or non-geometric compactifications, if they admit unrestricted fluxes unconstrained by tadpoles, can be engineered to yield scale-separated vacua with full moduli stabilization and controlled string corrections.

The following table summarizes several explicit constructions:

Type	Ingredients	Unbounded Flux	$L_{\text{AdS}}/L_{\text{KK}} \to \infty$?
IIA/O6 CY	$F_0,F_4,H_3$, O6	$F_4$	Yes [DGKT, (Carrasco et al., 2023)]
IIA/O6 T$^6$	$F_0,F_4,H_3$, O6	$F_4$	Yes (Apers et al., 2022, Shiu et al., 2022)
IIA G$_2$ O2/O6	$F_0,F_4,H_3$, O2/O6	$F_4$	Yes (Farakos et al., 2020, Miao et al., 16 Sep 2025)
IIB/O5/O7 SU(2)	$F_1,F_3,F_5,H_3$, O5/O7	$F_5$	Yes (Petrini et al., 2013)
Het G$_2$	$H_3$, gravitational instantons	$H_3$	Yes (Tringas et al., 11 Nov 2025)
AdS/CFT	Curvature-dominated, no orientifold	—	No

5. Constraints from the AdS Distance and Swampland Conjectures

The Strong AdS Distance Conjecture (SADC) posits that as $|\Lambda| \to 0$, a tower of states should become light with $m \sim |\Lambda|^{1/2}$. Generic classical compactifications without orientifolds or scale-separating fluxes obey $m_{\text{KK}} \sim |\Lambda|^{1/2}$, i.e., demonstrate no parametric scale separation (Gautason et al., 2015, Font et al., 2019).

In scale-separated orientifold vacua (e.g., DGKT), scaling the unbounded flux $N \to \infty$ with all moduli under control, one has: $$m_{\text{KK}} \ll m_{\text{AdS}},\quad \text{with}\quad m_{\text{KK}} \sim e^{-\alpha \Delta}$$ where $\Delta$ is a large geodesic distance traversed by an open-string modulus that interpolates between different large-$N$ vacua (Shiu et al., 2022). The KK tower becomes light along this path precisely in agreement with the Swampland Distance Conjecture (SDC), so these vacua evade SADC constraints when the strong SDC (in terms of field excursion) is satisfied.

More refined forms of the AdS Distance Conjecture require the presence of discrete higher-form symmetries and associated domain wall spectra; certain families of 3d scale-separated vacua (e.g., $AdS_3$ G$_2$-orientifold solutions) do not manifest the necessary discrete symmetry structure, and hence appear disfavored under these stronger conjectures (Apers et al., 2022, Hemelryck, 7 Feb 2025).

6. Robustness, Mass Spectra, and Extensions

Scale-separated AdS vacua of DGKT type and their T-dual/M-theory/heterotic uplifts display a striking universality:

• The lightest scalar operator dimensions (in AdS units) are integers, e.g., $\Delta \in \{5,6,10,11,\dots\}$ for DGKT-type solutions (Apers et al., 2022, Andriot et al., 2023). This integer spacing persists under subleading corrections (e.g., backreaction, warping, or higher large-flux expansions) as long as flux quantization is respected and all moduli remain stabilized.
• Perturbative spectra in these vacua are stable, with all tachyons (if present) lying above the Breitenlohner–Freedman bound. Twisted-sector moduli and non-perturbative membrane states may be included without destabilizing the vacuum (Carrasco et al., 2023, Andriot et al., 2023).
• Extensions including anisotropic scaling or more intricate metric/torsion structures (e.g., elliptic fibrations or K3, nilmanifold, solvmanifold internal spaces) yield broader families of vacua with controlled scale separation and stable spectra, though with non-universal gaps in exceptional cases (Carrasco et al., 2023, Tringas, 2023).

Flux-backtracking techniques have been used to relate the brane origin of these vacua: for instance, the near-horizon limit of inserting $N$ D4-branes into a DGKT-type singularity reproduces the AdS vacuum, giving a holographic interpretation even without a known explicit CFT dual (Apers et al., 3 Jun 2025).

7. Examples Beyond Standard Orientifold Constructions

A notable exception is the existence of scale-separated AdS$_3 \times S^3$ solutions in six-dimensional gauged supergravity (Salam–Sezgin model) without the need for orientifolds or exotic stringy sources. In the near-horizon limit of a rotating dyonic string, the Kaluza–Klein towers admit a limit where all but a finite set of low-lying states acquire arbitrarily large masses; in particular, only a finite set of states (with integer or half-integer conformal dimensions) remains light and all others decouple. This mechanism relies purely on 6D R-symmetry gauging and squashing, and differs qualitatively from the string compactification constructions (Proust et al., 16 Apr 2025).

Similarly, genuine scale-separated AdS$_3$ vacua have now been realized in the classical heterotic string, using only NSNS $H$-flux, gravitational instantons, and $G_2$-structure compactification; in these, large flux yields large volume, weak coupling, and $L_{\text{KK}} / L_{\text{AdS}} \to 0$, thus fully realizing scale separation in a corner where homogenous sphere or torus compactifications fail (Tringas et al., 11 Nov 2025).

---

• "On scale separation in type II AdS flux vacua" (Font et al., 2019)
• "Remarks on scale separation in flux vacua" (Gautason et al., 2015)
• "Supersymmetric AdS vacua and separation of scales" (Tsimpis, 2012)
• "AdS$_2$ Type-IIA Solutions and Scale Separation" (Lust et al., 2020)
• "AdS vacua with scale separation from IIB supergravity" (Petrini et al., 2013)
• "Comments on classical AdS flux vacua with scale separation" (Apers et al., 2022)
• "AdS scale separation and the distance conjecture" (Shiu et al., 2022)
• "Scale separation from O-planes" (Tringas et al., 21 Apr 2025)
• "No-scale and scale-separated flux vacua from IIA on G2 orientifolds" (Farakos et al., 2020)
• "Supersymmetric scale-separated AdS$_3$ orientifold vacua of type IIB" (Hemelryck, 7 Feb 2025)
• "T-dualities and scale-separated AdS$_3$ in type I" (Miao et al., 16 Sep 2025)
• "New families of scale separated vacua" (Carrasco et al., 2023)
• "Anisotropic scale-separated AdS$_4$ flux vacua" (Tringas, 2023)
• "Extensions of a scale-separated AdS$_4$ solution and their mass spectrum" (Andriot et al., 2023)
• "Effective Theories as Truncated Trans-Series and Scale Separated Compactifications" (Emelin, 2020)
• "Scale-separated AdS$_4$ vacua of IIA orientifolds and M-theory" (Cribiori et al., 2021)
• "Backtracking AdS flux vacua" (Apers et al., 3 Jun 2025)
• "Scale separation on AdS$_3\times S^3$ with and without supersymmetry" (Proust et al., 16 Apr 2025)
• "Classical scale-separated AdS$_3$ vacua in heterotic string theory" (Tringas et al., 11 Nov 2025)