CFTs with Large Gap from Barnes-Wall Lattice Orbifolds

We investigate orbifolds of lattice conformal field theories with the goal of constructing theories with large gap. We consider Barnes-Wall lattices, which are a family of lattices with no short vectors, and orbifold by an extraspecial 2-group of lattice automorphisms. To construct the orbifold CFT, we investigate the orbifold vertex operator algebra and its twisted modules. To obtain a holomorphic CFT, a certain anomaly 3-cocycle $ω$ needs to vanish; based on evidence we provide, we conjecture that it indeed does. Granting this conjecture, we construct a holomorphic CFT of central charge 128 with gap 4.

💡 Research Summary

The authors address the long‑standing problem of constructing two‑dimensional conformal field theories (CFTs) with a large spectral gap, motivated by applications in AdS₃/CFT₂ holography and the quest for extremal or near‑extremal theories. Their strategy mirrors the classic construction of the Monster vertex operator algebra (VOA) V♮, but replaces the Leech lattice with higher‑dimensional Barnes‑Wall lattices BW(m).

Key ingredients:

1. Barnes‑Wall lattices – For odd m, BW(m) is an even unimodular lattice of dimension 2m. Its shortest vectors have length squared 2⌊m/2⌋, which grows with m, ensuring that low‑weight lattice excitations are absent. The authors focus on m = 7, giving a 128‑dimensional lattice whose VOA V_{BW(7)} has character
   Tr q^{L₀‑c/24}=1+128 q+8384 q²+… .

2. Extraspecial 2‑group E(m) – The automorphism group of BW(m) contains a large extraspecial 2‑group E(m) of order 2^{2m+1}. For m = 7 the group is denoted E(7). The authors prove (Theorem 3.5) that the natural lattice automorphisms lift to genuine automorphisms of the VOA, so E(7) acts faithfully on V_{BW(7)}.

3. Orbifold by E(7) – Taking the fixed‑point sub‑VOA V^{E(7)} eliminates all states up to weight 3 except the Virasoro descendants L_{‑2}|0⟩ and L_{‑3}|0⟩. Consequently the smallest possible weight of a non‑Virasoro primary is 4, i.e. the “gap” of the orbifold is at least 4.

4. Twisted modules and vacuum anomalies – For each g∈E(7) there exists a unique g‑twisted module W_g. The authors compute the projective action of the centralizer C_{E(7)}(g) on W_g, showing that the associated 2‑cocycles c_g are trivial. Moreover, all twisted sectors have vacuum anomaly ≥4, guaranteeing that adding twisted modules will not re‑introduce low‑weight primaries.

5. Anomaly 3‑cocycle ω – A holomorphic extension of V^{E(7)} to a full holomorphic VOA exists only if the obstruction class ω∈H³(E(7),U(1)) vanishes. The paper supplies two independent pieces of evidence: (i) restriction of ω to every cyclic subgroup satisfies the “type‑0” (level‑matching) condition; (ii) the triviality of all c_g implies ω must be trivial (via the relation (1.3.1) from EG18). Although a full proof is not given, the evidence is compelling.

6. Holomorphic extension and final theory – Assuming ω=1, the modular tensor category of V^{E(7)} is the untwisted Drinfeld double D(E(7)). This permits a unique holomorphic extension V_{orb} with central charge c=128. Its character begins
   Tr q^{L₀‑c/24}=1+q⁴+O(q⁵),
   confirming a gap of 4.

Significance: This is the first known example of a two‑dimensional CFT (and holomorphic VOA) with central charge 128 whose smallest non‑vacuum primary has weight 4, surpassing all previously constructed lattice or permutation orbifolds that achieve at most gap 2. The result provides a concrete, sparsely populated spectrum suitable for testing holographic bounds, modular bootstrap constraints, and recent conjectures on large‑c CFT spectra (e.g.,