Mass Formula for Topological Boundaries

• The paper introduces a mass formula that rigorously quantifies topological boundary conditions using Lagrangian subgroups and modular tensor categories.
• It derives precise expressions for ground state degeneracy in topologically ordered phases, distinguishing models such as the toric code and double-semion.
• It connects algebraic classifications with physical boundary effects in TQFT, Dirac Hamiltonians, and higher-dimensional systems, providing unified insights.

A mass formula for topological boundary conditions provides a rigorous, quantitative characterization of the total (often weighted) set of admissible boundary conditions consistent with the bulk data of a theory, especially in the context of topological quantum field theories (TQFTs), topologically ordered condensed matter phases, and gauge/gravitational settings with nontrivial boundaries. The terminology is used in several distinct but mathematically related contexts: as the count of Lagrangian subgroups (classifying fully gapped boundary conditions) in modular tensor categories (TQFT), as the formula for ground state degeneracy (GSD) in topological phases as a function of boundary parameters, as the algebraic relation between Wilson-mass coefficients and topological boundary conditions in lattice Dirac models, and as the vacuum energy or topological mass term in quantum field theories with boundaries and external fields.

1. Mass Formula in Modular Tensor Categories and TQFT-Gravity

The mass formula in 3d TQFT, as formulated in "Mass formula for topological boundary conditions from TQFT gravity" (Dymarsky et al., 30 Jan 2026), gives the total weighted count of topological boundary conditions (TBCs). For an Abelian theory with bulk anyon group $\mathscr D$ and Hilbert space $\mathcal H^g$ on a genus-$g$ surface $\Sigma_g$, each TBC corresponds to a Lagrangian subgroup $\mathcal C \subset \mathscr D$. The mass is defined as

$M(\mathcal T) = \sum_{[\mathcal C]} 1$

in the Abelian case, and as

$$M(\mathcal T) = \sum_{[\mathcal C]} \frac{1}{|\mathrm{Aut}(\mathcal C)|}$$

in the non-Abelian case, where $\mathrm{Aut}(\mathcal C)$ is the automorphism group of the Lagrangian algebra.

A key result is the holographic interpretation:

$$M(\mathcal T) = \lim_{g\to\infty} \frac{D^g}{|\mathrm{MCG}(\Sigma_g)|} \sum_{\gamma \in \mathrm{MCG}(\Sigma_g)} \langle 0 | U_\gamma | 0 \rangle^g$$

where $D = \sqrt{|\mathscr D|}$ is the total quantum dimension, and $U_\gamma$ is the mapping-class group representation. This represents a genus-reduced, renormalized sum over partition functions on all closed 3-manifolds, identifying the mass as both an algebraic and topological gravity partition function normalization.

2. Mass Formula and Gapped Boundary GSD in Topological Orders

In 2+1D Abelian Chern-Simons theories, the mass formula manifests as the ground state degeneracy (GSD) on manifolds with boundary, dependent on the choice of Lagrangian subgroups (gapped boundary data) (Wang et al., 2012). For a $K$-matrix abelian theory, boundaries $\partial_\alpha$ are labeled by maximal, null sublattices $\Gamma^{\partial_\alpha}$ (interpreted as condensate lattices), and compatible quasiparticle lattices $\Gamma_{qp}^{\partial_\alpha}$. The general GSD formula for $\eta$ boundaries is

$$\mathrm{GSD} = \left| \left\{ (\ell^{(1)}_{qp}, \ldots, \ell^{(\eta)}_{qp}) \ \middle|\ \ell^{(\alpha)}_{qp} \in \Gamma_{qp}^{\partial_\alpha},\ \sum_{\alpha} \ell^{(\alpha)}_{qp} \in \Gamma_e \right\} \Big/ \left\{ (\ell^{(1)}, \ldots, \ell^{(\eta)})\ \middle|\ \ell^{(\alpha)} \in \Gamma^{\partial_\alpha} \right\} \right|$$

This GSD is finer than the bulk fusion algebra: it distinguishes topologically ordered phases with identical anyon content but different boundary degeneracies, as shown explicitly in the $Z_2$ toric code versus $Z_2$ double-semion model.

3. Algebraic Structure and Relation to Self-Dual Codes

For Abelian TQFTs (e.g., Dijkgraaf-Witten or toric code models), classification of TBCs is formally equivalent to classifying self-dual codes over finite fields or rings. Each Lagrangian subgroup corresponds to a maximal isotropic subspace or self-dual code. The mass formula in these cases reproduces classical results:

• For $(U(1)_p \times U(1)_{-p})^n$ (level-$p$ toric codes):

$M = \prod_{i=0}^{n-1}(p^i + 1)$

• For $U(1)_2^n$ (binary Type-II codes):

$M = \prod_{i=0}^{n/2 - 2}(2^i + 1)$

Factorization over prime powers and extension to higher $k$ involve arithmetic of symmetric forms and Gauss sums over $\mathbb Z_{p^m}$.

4. Dirac Hamiltonians, Wilson Mass, and Berry–Mondragon Boundary Conditions

In topological materials modeled by Dirac-like Hamiltonians, the "mass formula" links the Wilson mass $m$ to the realization of topological boundary conditions. For a 1D Dirac model $H(k) = \hbar v_F U_k k + m U_w k^2$, introducing the Wilson mass simultaneously regularizes the spectrum (avoiding fermion doublers) and enforces Dirichlet (Berry–Mondragon) boundary conditions (Araújo et al., 2019). Explicitly, the allowed range (on a lattice of spacing $\delta_x$) is

$$\frac{1}{2}\delta_x^2 \Delta_\epsilon \leq |m| \leq \frac{(\hbar v_F)^2}{\Delta_\epsilon}$$

where $\Delta_\epsilon$ is the low-energy window of interest. The Wilson term $H_W = m U_c k^2$ and the boundary condition matrix $M = (i U_k^{-1} U_c)/\alpha$ are directly related, ensuring physical matching at boundaries and eliminating spurious modes. This framework applies to graphene nanoribbons (yielding the generalized Brey–Fertig and Berry–Mondragon conditions) and topological crystalline insulators, and provides numerically stable lattice regularizations.

5. Vacuum Energy and Emergent Topological Mass from Boundary Conditions

In scalar QFTs with boundaries and background fields, topological boundary conditions induce a finite "topological mass", an effect that can dominate the low-energy spectrum (Junior et al., 20 Aug 2025). For a charged/neutral scalar pair on an interval, the total physical mass squared is

$$m_{\rm top}^2 = m^2 + \Delta m_{\rm boundary}^2 + \Delta m_{\rm magnetic}^2$$

where

$$\Delta m_{\rm boundary}^2 = \frac{\lambda_\psi m}{8\pi^2 L}\sum_{n=1}^\infty \frac{K_1(2 n m L)}{n}, \quad \Delta m_{\rm magnetic}^2 = \frac{g}{2\pi^2 L^2}\sum_{j=1}^\infty (-1)^j \mathcal I_1(j, L, 0)$$

and $K_1$ is the modified Bessel function, $\mathcal I_1$ encodes Landau quantization, $L$ is the separation of plates. In both strong and weak field regimes, the resulting topological mass exhibits universal exponential, polynomial, or logarithmic dependence on $B$ and $L$, controlled by the boundary and bulk data.

6. Five-Dimensional Generalizations and Higher-Dimensional Isotropy

The mass formula extends naturally to higher-dimensional Abelian TQFTs, notably 5d Abelian 2-form Chern-Simons ("BF") theories (Dymarsky et al., 30 Jan 2026). Here, the problem becomes one of classifying maximal isotropic (symplectic) subgroups of a finite group, mirroring the self-dual code analogy in higher dimensions. The mass formula for $g$ copies at level $k$ is, for $k$ prime,

$$\mathcal N\bigl((\text{2-form}_k)^{\otimes g}\bigr) = \prod_{i=1}^g (p^i + 1)$$

and similarly for higher $k$ in terms of antisymmetric matrix orbits.

7. Physical and Mathematical Implications

The mass formula for topological boundary conditions unifies algebraic, topological, and geometric criteria for counting gapped boundary types, linking TQFT classification, code theory, and explicit boundary-induced effects in QFT and condensed matter systems. In TQFT/gravity, it normalizes the partition sum over geometries and is conjecturally related to structures in string theory thresholds and modular forms. In topologically ordered phases, it provides operational invariants distinguishing theories with the same bulk data. In lattice Hamiltonians and quantum fields, it prescribes robust matching conditions controlling spectral and dynamical properties at boundaries. The mass formula thus serves as a fundamental organizing principle across a spectrum of mathematical physics domains (Wang et al., 2012, Araújo et al., 2019, Junior et al., 20 Aug 2025, Dymarsky et al., 30 Jan 2026).