O(16)$ imes$O(16) heterotic theory on $AdS_3 imes S^3 imes T^4$

In this paper, we study non-supersymmetric $AdS_{3}\times S^{3}$ vacuua of $O(16)\times O(16)$ heterotic theory on a string scale $T^{4}$ background, which are parameterized by a pair of flux integers. Adding the one-loop scalar potential to the effective theory contributes positively to the cosmological constant, but we find that there is no uplift to de Sitter for any values of the fluxes. We study the fluctuations around these vacua and show that all scalar and tensor modes from the six-dimensional effective theory lie above the Breitenlohner-Freedman bound. The moduli coming from the torus compactification will also be above the bound, at least for a large range of fluxes.

💡 Research Summary

The paper investigates non‑supersymmetric vacua of the ten‑dimensional O(16)×O(16) heterotic string compactified on a string‑scale four‑torus (T⁴) and further on an AdS₃×S³ background. The authors parameterize the solutions by two integer fluxes: n₅, the magnetic H₃ flux threading the three‑sphere, and n₁, the electric H₇ flux (dual to H₃) threading the product S³×T⁴.

At tree level the effective six‑dimensional action contains contributions from the curvature of the internal S³, the magnetic flux term proportional to n₅², and the electric flux term proportional to n₁². Solving the extremisation conditions for the dilaton (ϕ) and the S³ radius modulus (χ) with ϕ=χ=0 yields a simple relation: the AdS₃ radius squared L² = α′ n₅ and the string coupling gₛ² = v n₅ (2π)⁴ n₁⁻¹, where v is the dimensionless volume of the torus. The resulting three‑dimensional cosmological constant is Λ₃ = –1/(α′ n₅), i.e. a negative AdS value.

The novel part of the work is the inclusion of the one‑loop correction to the scalar potential. This correction is obtained by integrating the genus‑one partition function of the O(16)×O(16) heterotic string over the fundamental domain of the modular parameter τ. The authors define a dimensionless coefficient λ = –π² v F_d²(τ) Z(τ), which they evaluate for a square torus at the self‑dual radius and find λ≈1.321. The one‑loop contribution to the three‑dimensional potential takes the form V₁‑loop = λ gₛ² α′⁻¹ e^{–3χ}. Adding this to the tree‑level potential yields a total potential V(ϕ,χ) whose stationary equations become a coupled set of algebraic equations. By eliminating gₛ² e^{–6χ} the authors reduce the problem to a quartic equation for L⁴, which they solve analytically. The solutions are expressed in terms of n₁, n₅ and λ, and are plotted in Figure 1. The plots show that for all admissible flux values the corrected vacuum energy V_min remains negative; the one‑loop term never uplifts the vacuum to a de Sitter (positive) cosmological constant. This result is consistent with existing no‑go theorems, and the authors argue that even when electric flux is present the same obstruction persists.

Stability is examined by expanding fluctuations of the six‑dimensional supergravity fields (scalars, tensors) in spherical harmonics on S³. The mass‑squared spectrum is computed and compared with the Breitenlohner‑Freedman (BF) bound for AdS₃, m²_{BF}=–(d²)/(4ℓ²) with d=2. All modes derived from the six‑dimensional sector lie above this bound. The torus moduli—volume, complex structure, and B‑field—also acquire masses from the fluxes and the one‑loop potential. For a broad range of flux choices these moduli remain safely above the BF bound; only in the strict n₁→0 limit do some masses approach the bound, but the one‑loop correction prevents any violation.

In summary, the O(16)×O(16) heterotic theory admits a family of non‑supersymmetric, perturbatively stable AdS₃×S³×T⁴ vacua parametrized by (n₁,n₅). The one‑loop scalar potential, while positive and capable of stabilising the dilaton, is insufficient to generate a de Sitter uplift. All scalar and tensor fluctuations satisfy the BF stability criterion, and the torus moduli are likewise stable for a large flux window. The work provides a concrete example of how quantum corrections affect non‑supersymmetric string compactifications, confirming that de Sitter uplift remains elusive in this setting and highlighting the robustness of AdS stability even without spacetime supersymmetry. Future directions suggested include higher‑loop corrections, alternative torus shapes, or additional fluxes to explore possible routes toward positive vacuum energy.