Geometric realization of algebraic conformal field theories (Q2419272)

In the paper under review, the author presents a geometric perspective on the relationship between two formulations of the two dimensional chiral conformal field theory, i.e., the vertex operator algebra (VOA) and the conformal nets, initiated in [\textit{S. Carpi} et al., From vertex operator algebras to conformal nets and back. Providence, RI: American Mathematical Society (AMS) (2018; Zbl 1434.17001)], based on a geometric picture of conformal nets introduced in [\textit{A. Henriques}, Contemp. Math. 613, 1--40 (2014; Zbl 1322.81061)], and a rigorous construction of the unitary free fermion theory [\textit{J. E. Tener}, Commun. Math. Phys. 355, No. 2, 463--518 (2017; Zbl 1406.81085)] which materializes the Segal's geometric definition of a conformal field theory in terms of functors from the two dimensional complex bordism category. \par To continue we need the following three pieces of information: (1) In the Segal's picture, there is a map \(T: \mathcal{H}\otimes \mathcal{H}\to\mathcal{H}\) assigned to a disk with two disks removed where \(\mathcal{H}\) is a Hilbert space assigned to the circle \(S^1\). (2) The principal piece of data for a conformal net is a family of von~Neumann algebras $\mathcal{A}(I)$, called the local algebras, indexed by intervals $I \subset S^1$. (3) In Henriques' geometric perspective, the local operators of a conformal net correspond to degenerate annuli with zero thickness along $I^c$, with states inserted in the thick part of the annulus. \par Considering the degenerate annulus as the limit $\lim_{R\downarrow1} X_{R,t}$ of a family of Riemann surfaces $X_{R,t}:=R\mathbb{D}\setminus(\phi_t(\mathring{\mathbb{D}})\cup (w+s\mathring{\mathbb{D}}))$, where $(\phi_t)_{t\ge0}$ is a one-parameter semigroup of holomorphic injective self maps of the unit disk $\mathbb{D}$ (whose interior is $\mathring{\mathbb{D}}$) which fix 0 and map onto Jordan domains with $C^\infty$ boundary, the author proves that there exists a map $T_{R,t}:\mathcal{F}\otimes\mathcal{F}\to\mathcal{F}$ corresponding to each $X_{R,t}$, where $\mathcal{F}$ denotes the fermionic Fock Hilbert space, and the limit $\lim_{R\downarrow1} T_{R,t}$ converges to a {\em bounded} operator $T_{t}:\mathcal{F}\otimes\mathcal{F}\to\mathcal{F}$ in the strong operator topology. Moreover, $T_t$ can be characterized in terms of commutation relations with generators of the ``canonical anticommutation relations'' algebra determined by the Hardy space of the degenerate annulus. \par Furthermore the author proves the following theorem: Let $\mathcal{F}^0$ be the subspace of the finite energy vectors of the free fermion VOA, regarded as a subspace of its Hilbert space completion $\mathcal{F}$. Let $V \subset (\mathcal{F}^0)^{\otimes N}$ be a unitary vertex operator subalgebra, and for \(a \in V\) let $T_{t;a}(\cdot):=T_t(a\otimes\cdot)$. Then $T_{t;a}$ is {\em bounded}, and \[ \mathcal{A}_V (I) :=\{U(\gamma_1)^*T_{t;a}U(\gamma_2), (U(\gamma_1)^* T_{t;a}U(\gamma_2))^*\mid a \in V, (\gamma_1, \gamma_2) \in \mathscr{P}_I\}'', \] defines a conformal net on the Hilbert space completion of $V$, with conformal symmetry $U$ given by the positive energy representation of $\text{Diff}_+(S^1)$ induced by the conformal vector of $V$. Here, $\text{Diff}_+(S^1)$ is the group of orientation preserving diffeomorphisms of the unit circle, $\mathcal{A}''$ denotes the bicommutant of $\mathcal{A}$, and $\mathscr{P}_I$ is the collection of all pairs $(\gamma_1, \gamma_2)$ such that \(\gamma_1(z) = \phi_t(\gamma_2(z))\) for all $z$ lying in some open, connected, non-empty, non-dense subset $I \subset S^1$. \par The author provides many examples of VOAs which appear as subalgebras of $N$ free fermions, for some $N$, including the free boson, lattice VOAs corresponding to sublattices of $\mathbb{Z}^N$, the discrete series of Virasoro VOAs and affine VOAs corresponding to simple Lie algebras $\mathfrak{g}$, at levels $k\Delta_g$ $(k = 1, 2, \cdots)$, where $\Delta_g \in \mathbb{Z}_{>0}$ is a constant depending on $\mathfrak{g}$.