Synthetic Hall Torus: Engineered Quantum Topology

• Synthetic Hall torus is an engineered quantum system with a torus geometry, featuring periodic real and synthetic dimensions intertwined by gauge-invariant flux.
• It employs various experimental setups such as ring traps and optical lattices to realize quantized Hall responses and controlled topological transport.
• The system leverages Raman-induced tunneling and synthetic dimensions to explore quantum Hall physics, band topology, and interaction-driven many-body phenomena.

Synthetic Hall torus denotes an engineered quantum system whose effective two-dimensional configuration space is a torus and whose dynamics are governed by gauge-invariant Peierls phases or boundary twists that emulate magnetic flux. In the literature, the most common realizations combine a periodic real coordinate—such as a ring angle or a Bloch-periodic lattice direction—with a periodically closed synthetic dimension formed from internal states, but real-space torus optical lattices and phase-space tori of quasiperiodically driven qubits are also described in the same language. The unifying structure is not a particular apparatus but the conjunction of toroidal boundary conditions, synthetic gauge flux, and Hall-type topological response [1612.00233, 1805.01483, 2602.14549, 2004.03457].

1. Geometry and defining constructions

In the synthetic-dimension formulation, one coordinate is a real spatial variable and the second coordinate is an internal-state index. A representative construction uses one continuous real-space coordinate (x) together with a synthetic dimension (y=1,\ldots,M) built from hyperfine spin states, where Raman transitions couple (y \to y+1) with a Peierls phase (e^{i b x}). When the synthetic dimension is closed by cyclic couplings, (y=M+1\equiv 1), both directions are periodic and the geometry is a torus; with open synthetic ends the geometry is a cylinder [1612.00233]. In related ring-based realizations, the real coordinate is the azimuthal angle (\theta) of a ring trap, while the synthetic direction is a loop of spin states (s=1\to2\to3\to1), again producing a torus when both cycles are closed [2602.14549].

The same topological objective can be achieved without using a synthetic dimension as the second coordinate. A bilayer square optical lattice with inter-layer tunneling activated only along the edge of a square annulus closes the tight-binding network into a torus with two non-contractible cycles (a) and (b). In that scheme, curvature is encoded in connectivity rather than in a physically bent surface [1805.01483]. A further generalization treats the pair of drive phases ((\theta_1,\theta_2)) of a quasiperiodically driven qubit as coordinates on (T^2=[0,2\pi)\times[0,2\pi)), so that the “torus” is a synthetic phase space rather than a spatial manifold [2004.03457].

This multiplicity of constructions makes one recurring distinction essential: a Hall cylinder is not a Hall torus. In the continuous-(x) synthetic-dimension setting, open synthetic boundaries allow the Peierls phase to be gauged away within a single atom’s spin manifold, yielding effectively a one-dimensional gas, whereas periodic closure of the synthetic direction makes the phase winding gauge-invariant and generates the Hall-torus configuration central to the toroidal phenomena [1612.00233]. A closely related statement appears in the four-state Hall-cylinder realization: with open synthetic boundary the Raman phases can be gauged away and the characteristic band crossings disappear, while closing the synthetic loop restores the nontrivial topology [1809.02122].

Realization class	Periodic directions	Characteristic consequence
Continuous (x) + synthetic spin loop	(x) and synthetic dimension	Uniform flux and thin-torus-like crystalline states
Ring BEC + cyclic spin couplings	Azimuthal angle (\theta) and spin loop	Quantized toroidal flux and azimuthal density modulation
Bilayer optical lattice torus	Two real-space cycles	Quantized supercurrents and torus FQH degeneracy

2. Gauge structure and microscopic Hamiltonians

The canonical microscopic mechanism is Raman-induced complex tunneling. In the continuous-(x), (M)-component setting, a representative single-particle Hamiltonian is
$$
H_0 = \int_0^{L_x} dx\, \sum_{m=1}^{M} \Psi_m^\dagger(x)\left[-\frac{\hbar^2}{2M_a}\,\partial_x^2 + \Delta_m\right]\Psi_m(x) -\Omega \int_0^{L_x} dx\, \sum_{m=1}^{M} \left[e^{i b x}\,\Psi_{m+1}^\dagger(x)\Psi_m(x) + \text{h.c.}\right],
$$
with periodic boundary conditions (\Psi_{M+1}(x)=\Psi_1(x)) for the torus. The spatially varying phase produces a uniform flux (\phi=b\Delta x) through each elementary (x)–synthetic plaquette. In the Wannier-orbital description, the natural lattice spacing is (a=2\pi/(Mb)), so a representative plaquette carries (\phi=b a=2\pi/M), and over a length (L_x) satisfying (bL_x=2\pi N_x) the total number of flux quanta is (N_\phi=N_xM) [1612.00233].

In ring-based realizations the gauge structure is encoded in orbital angular momentum transfer rather than linear momentum transfer. For a spinor Bose–Einstein condensate on a ring, two Raman links carry the position-dependent phase (e^{i\theta}) and the microwave closing link carries a uniform phase (e^{-i\theta_{\mathrm{mw}}}). Writing the couplings as (H_{\mathrm{cpl}}(\theta)=\sum_{s=1}^3[\Omega_s e^{i\phi_s(\theta)}|s+1\rangle\langle s|+\mathrm{h.c.}]) with ((\phi_1,\phi_2,\phi_3)=(\theta,\theta,-\theta_{\mathrm{mw}})), the gauge-invariant loop phase is
$$
\Phi_{\mathrm{loop}}(\theta)=2\theta-\theta_{\mathrm{mw}}.
$$
As (\theta\to\theta+2\pi), this yields (\Delta\Phi_{\mathrm{loop}}=4\pi), i.e. two flux quanta threaded through the toroidal surface, independent of (\theta_{\mathrm{mw}}) [2602.14549].

A more general torus construction based on Laguerre–Gaussian Raman beams expresses the total synthetic flux through the torus in terms of the accumulated momentum transfers along the synthetic loop. If each link (j\to j+1) carries phase (e^{i q_{j,j+1}x}), then
$$
Q=\sum_{j=1}^N q_{j,j+1}, \qquad W_{\mathrm{synth}}(x)=e^{iQx},
$$
and the total flux through the whole torus is (\Phi_{\mathrm{torus}}=QL/(2\pi)). Because the gauge field lives partly in a synthetic coordinate, this net effective flux through a torus surface is not restricted by (\nabla!\cdot!B=0) in ordinary three-dimensional space [1810.12331].

3. Effective low-energy theories and many-body structure

For large Raman coupling in the continuous-(x) synthetic-dimension model, Fourier transformation along the synthetic direction yields independent sectors
$$
h_{k_y} = -\frac{\hbar^2}{2M_a}\partial_x^2 - 2\Omega \cos(bx-k_y).
$$
In the deep-dressing regime (4\Omega\gg E_r), these sectors form deep cosine potentials whose minima are evenly spaced by (a=2\pi/(Mb)). Projecting onto localized Wannier orbitals produces an effective one-dimensional lattice with hopping by exactly (M) sites and long-ranged density-density interactions,
$$
H_{\mathrm{eff}} = -t \sum_{j}\left(a_{j+M}^\dagger a_j + \text{h.c.}\right) + \sum_{r\ge 0} V_r \sum_j :n_j n_{j+r}: - \mu \sum_j n_j,
$$
where (V_r = \frac{g}{\sqrt{2\pi}\,\ell} e^{-(a^2/2\ell^2)r^2}). In the large-(M) limit the hopping is exponentially suppressed and the model becomes interaction-dominated, closely paralleling the thin-torus limit of quantum Hall systems [1612.00233].

In the convex regime, (V_{r+1}+V_{r-1}>2V_r), fermions or hard-core bosons exhibit a complete devil’s staircase of crystal ground states at every rational density (\rho=p/q), equivalently at every rational filling (\nu=p/q) because (\rho=N/N_\phi=\nu). Convexity for the Gaussian interaction holds when (\ell/a<\sigma_0\approx1.957). The crystal at (\rho=1/q) is the most homogeneous pattern (\overline{q}), while at general (\rho=p/q) the unit cell has length (q) with separations as uniform as possible. Plateau widths decrease monotonically with (q), so low-denominator states are widest. For finite (M), hopping melts higher-denominator crystals more easily; the analysis finds (V_0\gg V_1\gg J\gg V_2) for (M=2,3), stabilizing (\rho=1) and (\rho=1/2), while for (M\ge4) the (\rho=1/3) crystal survives and larger (M) stabilizes larger denominators [1612.00233].

The relation to torus fractional quantum Hall theory is explicit in (K)-matrix formulations. For the multilayer Abelian torus model specified by a complex torus (E=\mathbb{C}/(\mathbb{Z}+\tau\mathbb{Z})) and a symmetric positive definite integer matrix (K), the many-body ground-state space has dimension
$$
\delta=\det K.
$$
The corresponding wave functions form a holomorphic vector bundle over the torus of Aharonov–Bohm phases, with first Chern class
$$
c_1(\mathcal{E}) = -\frac{i}{2t}\sum_{p,q=1}^g K^\sharp_{pq}\, d\xi_p \wedge d\bar{\xi}_q,
$$
and the center-of-mass Hermitian metric yields a projectively flat Bott–Chern connection [2309.04866]. This places the synthetic Hall torus within the same mathematical framework as torus ground-state degeneracy, magnetic translations, and Berry curvature in conventional quantum Hall theory.

4. Symmetry, band topology, and pumping

Synthetic Hall tori support symmetry structures absent in planar or open-boundary counterparts. In the ring-plus-synthetic-loop construction with commensurate momentum transfers (q_{m,m+1}=n_m q_L), the Hamiltonian is periodic with period (d=2\pi/q_L), yet the density has the shorter period
$$
\lambda_\rho = \frac{2\pi}{Q} = \frac{d}{n}, \qquad n=\frac{Q}{q_L}.
$$
This fractionalization is tied to a nonsymmorphic symmetry (G) combining translation by (2\pi/Q) with an internal-state unitary. Since (G^n=T(d)), the symmetry sectors permute under (k\to k+q_L), forcing band connectivity and protected crossings within each (n)-band cluster. Under a constant force, wavepackets therefore braid across bands and return to themselves only after (\Delta k=n q_L), rather than after one Brillouin zone [1810.12331].

A closely related toroidal pumping structure appears in the three-leg synthetic Hall tube with tunable threaded flux (\Phi). In the gauge (\phi_{j;21}=\phi_{j;32}=j\phi_0), (\phi_{j;13}=j\phi_0+\Phi), with (\phi_0=2\pi/3), the Bloch Hamiltonian depends periodically on ((k,\Phi)), and the Chern number of band (n) is
$$
C_n = \frac{i}{2\pi}\int dk\, d\Phi \left[\langle \partial_\Phi u_n|\partial_k u_n\rangle - \langle \partial_k u_n|\partial_\Phi u_n\rangle\right].
$$
Adiabatically cycling (\Phi) by (2\pi) implements a Laughlin–Thouless pump with quantized transported charge (\Delta Q=C). When the real direction is made periodic, the tube becomes a torus: edge states disappear, but the Chern numbers and pump quantization remain [2002.07617].

The four-state Hall-cylinder realization exposes another toroidal symmetry mechanism. There, a nonsymmorphic symmetry (\hat{S}=\hat{U}\hat{G}), with (\hat{G}=e^{-i\hat{p}_y d/(2\hbar)}) and (\hat{U}=\mathrm{diag}(1,-1,-1,1)), protects band crossings and produces a period multiplier (\mu=2) in Bloch oscillations. Adding radio-frequency couplings breaks the symmetry and opens a gap; an axial phase (\theta_{\mathrm{axial}}=2\theta_{\mathrm{RF}}+\theta_1-\theta_2) then acts as a tunable axial flux. The same paper argues that imposing periodic boundary conditions along the real direction closes the cylinder into a torus while preserving the symmetry-protected crossings and Möbius-like transport when the circumference is commensurate [1809.02122].

5. Experimental realizations and diagnostics

The first experimental realization explicitly identified as a synthetic Hall torus used a spinor (^{87}\mathrm{Rb}) Bose–Einstein condensate confined in a ring-shaped trap of radius (R\approx14\,\mu\mathrm{m}). The synthetic dimension was formed by the three (F=1) Zeeman sublevels (m_F=+1,0,-1), cyclically coupled by Raman and microwave fields. The Raman configuration used a Gaussian beam and a Laguerre–Gaussian beam with winding number (1) at (\lambda=790\,\mathrm{nm}), with (\Omega_M/2\pi=3.8\,\mathrm{kHz}) and (r_M=17\,\mu\mathrm{m}); the microwave closing link had (\Omega_{\mathrm{mw}}/2\pi=3.15\pm0.25\,\mathrm{kHz}). With the synthetic dimension open, no two-node structure appeared. With the microwave link turned on, the condensate developed two azimuthal maxima separated by approximately (\pi), with (\mu_1=(0.62\pm0.02)\pi), (\mu_2=(1.69\pm0.02)\pi), widths near (0.14\pi), and a phase-response slope (d\mu/d\theta_{\mathrm{mw}}\approx0.51). The modulation emerged within approximately (6\,\mathrm{ms}), and abrupt quenches produced width oscillations at approximately (39)–(41\,\mathrm{Hz}) [2602.14549].

A complementary route constructs a real-space torus in an optical lattice. In the bilayer square-annulus design for (^{87}\mathrm{Rb}), the horizontal lattice spacing is (a_x\simeq a_y=480\,\mathrm{nm}), and region-dependent vertical potentials are tuned so that inter-layer tunneling is enabled only along the annulus edge. Numerical estimates give intra-layer bulk-bulk tunneling (\approx0.032\,E_r), intra-layer edge-bulk (\approx0.041\,E_r), inter-layer edge-edge (\approx0.036\,E_r), and inter-layer bulk-bulk (\approx0.002\,E_r). On this torus, weakly interacting condensates exhibit quantized supercurrents along the two non-contractible cycles, while exact diagonalization for (3) hardcore bosons on a (6\times6) torus at flux (\phi=\pi/3) shows the expected two-fold ground-state degeneracy at (\nu=1/2). The degeneracy persists for inter-layer tunneling down to approximately (0.7J), and under disorder of scale (0.05J) the splitting remains approximately (5\times10^{-3}J), much smaller than both the disorder scale and the excitation gap [1805.01483].

Interacting fermionic tubes provide an experimentally motivated intermediate system whose torus implications are explicit. For (^{173}\mathrm{Yb}), three hyperfine states are cyclically coupled to form a synthetic Hall tube with flux (\phi=2\pi/3). Density-matrix renormalization-group calculations at half-filling identify four gapped paramagnetic phases—NTSV, NTST, TSM, and NTSM—using entanglement spectra, entanglement entropy, chemical-potential spectra, and local spin-vector and spin-tensor observables. The paper emphasizes that imposing periodic boundary conditions in the real direction converts the tube into a torus without changing the bulk phase boundaries (\Omega_\pm=\pm3t+\sqrt{\Omega_{12}^2+9t^2}) or the interaction-driven transition near (U_c/t\approx1.85) for (\Omega_{31}/t=11) and (\Omega_{12}/t=12.3); what changes is the loss of edge diagnostics and the availability of torus-defined invariants such as many-body Chern numbers under boundary twists [2008.00708].

Across these implementations, the main diagnostics recur with different microscopic readouts. Density modulations and Bragg peaks track toroidal flux and crystalline order; spectral flow under boundary twists probes topological degeneracy; center-of-mass shifts and phase-controlled rotation of density extrema realize pumping; and quantized vorticities along the two non-contractible cycles diagnose toroidal superflow [1805.01483, 2602.14549].

6. Theoretical generalizations and conceptual scope

The term “synthetic Hall torus” also appears in settings where the torus is a parameter manifold rather than a spatial one. In a quasiperiodically driven nitrogen-vacancy-center qubit, the two drive phases satisfy (\theta_i(t)=\Omega_i t+\theta_{0i}) and each angle is defined modulo (2\pi), so ((\theta_1,\theta_2)) live on (T^2). Replacing crystal momentum ((k_x,k_y)) in the half-BHZ Hamiltonian by ((\theta_1,\theta_2)) yields a Chern-insulator structure on this phase torus, with Berry curvature
$$
F_{12}(\theta)=\partial_{\theta_1}A_2-\partial_{\theta_2}A_1,\qquad
C=\frac{1}{2\pi}\int_0^{2\pi}d\theta_1\int_0^{2\pi}d\theta_2\,F_{12}(\theta),
$$
and an adiabatic energy current between drives
$$
P_{1\to2}=\frac{\Omega_1\Omega_2 C}{2\pi}.
$$
Experimentally, overlap oscillations yielded (|C_{\mathrm{exp}}|=0.97\pm0.03) at (m=1), (|C_{\mathrm{exp}}|\le0.042) at (m=3), and (|C_{\mathrm{exp}}|=0.50\pm0.02) at the transition (m=2) [2004.03457].

At the opposite extreme of realism, a genuine toroidal surface can host a Hall response without synthetic dimensions. In the three-dimensional second-order topological-insulator proposal on a torus, a Zeeman term alone generates chiral hinge channels on the outer and inner equators. The boundary charge obeys
$$
Q_B^{R,L}(f)=\mp f \pm \Theta(f-f_{R,L}^*),
$$
leading to a quantized Hall conductance (\sigma_{\rho\phi}^{R,L}=e^2/h) under flux ramping. The quantization is unaffected by orbital Landau-level physics and survives disorder and smooth torus deformations [2212.09053]. This usage is conceptually adjacent rather than identical, but it reinforces the broader meaning of a Hall torus as a toroidal system with flux-driven Hall transport.

The torus geometry also imposes stringent wave-function constraints. In the lowest Landau level on a torus, projected delta-function coherent states provide a reproducing kernel and are maximally localized, whereas Haldane–Rezayi zero-locus states can develop a characteristic two-peak structure near (\mathrm{Re}\,\tau\approx1/2). Modular covariance restricts admissible torus trial states, and for the (\nu=2/5) problem it motivates replacing ordinary holomorphic derivatives by many-body magnetic translations with coefficients fixed by (S) and (T) transformations [1401.6834]. A plausible implication is that synthetic Hall tori are not only platforms for implementing toroidal topology, but also stringent laboratories for testing torus-specific localization, modularity, and Berry-connection structure.

Taken together, these works show that “synthetic Hall torus” is best understood as a topological design principle. The torus may be assembled from a real circle and a synthetic loop, from two stitched real-space cycles, or from periodic control phases; the flux may appear as a Peierls phase, an Aharonov–Bohm twist, or a Berry-curvature texture. What persists across the literature is the central triad of periodicity in two directions, gauge-invariant flux on the resulting torus, and Hall-type response encoded in crystalline order, band topology, pumping, degeneracy, or projectively flat Berry transport [1612.00233, 1805.01483, 2309.04866].