Momentum-Space Superradiance Lattices
- Momentum-space superradiance lattices are synthetic tight-binding models where timed Dicke states form discrete momentum sites, reversing the usual roles of real and reciprocal space.
- Coherent optical couplings, via standing-wave and Floquet modulation, engineer controlled tunneling, complex phase factors, and topological band structures in these lattices.
- Experimental platforms, including Bragg-coupled momentum-state systems and EIT configurations, reveal key phenomena like dynamic localization, Bloch oscillations, and chiral edge currents.
Momentum-space superradiance lattices are tight-binding lattices whose sites live in momentum space rather than real space. In the standard formulation, the sites are timed Dicke states of an ensemble of three-level atoms, and coherent optical couplings shift the collective excitation momentum by discrete amounts set by the coupling-field wave vectors, thereby producing a synthetic lattice for collective atomic excitations in momentum space (Wang et al., 2014). The concept has subsequently been extended to topological honeycomb lattices with Haldane-type band structure (Wang et al., 2015), Floquet-modulated momentum-frequency lattices in thermal atoms (Xu et al., 2022), and one-dimensional bipartite lattices with an exactly flat band and winding number $2$ (Li et al., 2023). Related Bragg-coupled momentum-state lattices, although not always formulated in explicitly superradiant language, have provided a parallel experimental and theoretical toolkit for programmable tunneling, disorder, topology, and spectroscopy in synthetic momentum dimensions (An et al., 2017).
1. Definition and conceptual basis
The defining objects of a superradiance lattice are timed Dicke states. For an ensemble of atoms at positions , absorption of a probe photon with wave vector prepares the collective state
whose spatial phase records the photon momentum and whose emission is superradiant in the direction (Wang et al., 2014). Introducing a metastable level and coupling fields with wave vectors and makes the collective momentum label dynamical: a coupling photon changes the timed Dicke momentum by , so the states 0 and 1 form a one-dimensional bipartite chain in momentum space (Wang et al., 2014).
This construction reverses the usual relation between real-space and reciprocal-space variables. In an ordinary crystal, position is discrete and quasimomentum is the Bloch label; in a superradiance lattice, the discrete label is the collective momentum, while a reciprocal coordinate 2 plays the role ordinarily played by quasimomentum. For the one-dimensional standing-wave construction with 3, the resulting bands are
4
so the lattice dynamics of timed Dicke excitations is mathematically identical to tight-binding dynamics, but in a synthetic momentum coordinate (Wang et al., 2014).
A closely related, experimentally mature notion is the momentum-state lattice built from discrete matter-wave momentum orders 5 with 6, coupled by two-photon Bragg transitions. In that setting, spectroscopic addressability of distinct Bragg resonances yields site-resolved control of tunneling amplitudes, phases, and on-site energies in a synthetic momentum-space chain (An et al., 2017). This broader synthetic-dimension framework has become an important methodological reference for momentum-space superradiance lattices.
2. Microscopic constructions and effective Hamiltonians
The original one-dimensional superradiance lattice is obtained in a 7-type EIT configuration with a weak probe on 8 and a standing-wave coupling field on 9. In the classical-field limit, the interaction Hamiltonian reduces to
0
which is a nearest-neighbor tight-binding model on a bipartite momentum-space chain (Wang et al., 2014). The two sublattices are the 1-type and 2-type timed Dicke states, and the detuning difference between the two standing-wave components provides a uniform force in momentum space rather than a static lattice tilt in real space.
A second major construction uses three EIT coupling fields with wave vectors at 3 to generate a honeycomb lattice of timed Dicke states in momentum space. Periodic modulation of the three coupling amplitudes,
4
produces, after Floquet elimination of the fast drive, complex next-nearest-neighbor hoppings
5
so the effective momentum-space model is a Haldane Hamiltonian with real nearest-neighbor hoppings, purely imaginary next-nearest-neighbor hoppings, and a sublattice offset 6 (Wang et al., 2015). In this form, the superradiance lattice becomes a topological Chern insulator for collective light-matter excitations.
A third construction replaces the three-level 7 scheme by a five-level M-type atomic ensemble. After adiabatic elimination of a far-detuned excited state, the timed Dicke basis 8 yields an effective one-dimensional bipartite lattice
9
with all 0 and 1 set by Rabi frequencies and detunings (Li et al., 2023). Under the symmetry conditions 2 and 3, one band is exactly flat, 4, while the other remains dispersive (Li et al., 2023).
| Construction | Site basis | Effective couplings |
|---|---|---|
| Standing-wave EIT chain | 5 | NN bipartite hopping 6 |
| Floquet honeycomb SL | timed Dicke honeycomb sites | NN 7, complex NNN 8 |
| Five-level M-type SL | 9 | 0, including extended SSH terms |
These constructions share a common structural principle: optical fields do not merely dress a real-space lattice; they directly define the graph, the hopping amplitudes, and the on-site structure of a lattice whose coordinate is collective momentum.
3. Momentum-space transport, localization, and coherent dynamics
In the one-dimensional EIT superradiance lattice, the detuning difference 1 between the two counter-propagating coupling components acts as a uniform momentum-space force,
2
This immediately gives the momentum-space analogue of several standard lattice phenomena. The Bloch-oscillation period is
3
the static-force problem generates Wannier-Stark ladders, and periodic modulation of the force yields Floquet quasienergies
4
so the effective bandwidth collapses when 5, producing dynamic localization (Wang et al., 2014). In this framework, Bloch oscillations are oscillations of the collective excitation over timed Dicke momentum sites, while the observable consequence is oscillatory superradiant emission and diffraction.
A related but conceptually complementary route to momentum-space lattice physics is furnished by the Aubry–André duality. For atoms in a one-dimensional incommensurate lattice,
6
the Fourier amplitudes 7 obey a tight-binding equation in momentum space. Localization occurs in momentum space for 8, and in that regime a three-mode approximation captures coherent oscillations between the central and side peaks, with frequency
9
Because this frequency is independent of the random lattice phase 0, phase-averaged oscillations survive in the momentum-localized regime and disappear beyond the transition (Larcher et al., 2010). For momentum-space superradiance lattices, this establishes a general principle: few-mode localization in momentum space and robustness against phase averaging are compatible.
Bragg-coupled momentum-state lattices have made the transport phenomenology of synthetic momentum-space chains quantitatively explicit. In a 21-site chain of 1 couplings, static random tunneling phases are dynamically irrelevant in one dimension and give ballistic spreading with variance exponent 2, while dynamical phase disorder produces nearly diffusive transport with 3, and Aubry–André site disorder yields arrested transport with 4 at 5 together with exponential localization 6 and 7 lattice sites (An et al., 2017). Although these experiments are not formulated as superradiance lattices, they demonstrate how coherent, diffusive, and localized transport regimes emerge in the same momentum-space tight-binding architecture.
4. Topology, Floquet gauge fields, and flat bands
The Floquet honeycomb superradiance lattice realizes a momentum-space Haldane model. In the effective two-band description, the masses at the two inequivalent Dirac points are
8
with
9
The Chern number is
0
and the topological transition occurs when
1
Experimentally accessible topology is encoded in the superradiant diffraction contrast
2
whose sign tracks the direction of the chiral excitation current in the superradiance lattice (Wang et al., 2015).
The five-level M-type construction realizes a one-dimensional extended SSH-type superradiance lattice with Bloch Hamiltonian
3
Under 4 and 5, the 6 term vanishes, one band remains exactly flat at 7, and the winding number
8
changes from 9 to 0 as the ratio 1 is varied through the gap-closing point (Li et al., 2023). The flat band itself carries a Berry-phase-based invariant 2, so the model realizes a topological flat band with higher winding number rather than a standard SSH chain.
Floquet superradiance lattices in thermal atoms extend the synthetic dimension by adding a frequency coordinate. With timed Dicke states 3 and 4, phase modulation
5
builds a two-dimensional lattice in momentum and Floquet index. Doppler shifts provide a force 6 along the momentum dimension, while Floquet replicas give 7, so the total force is 8. Transport along the crystal direction 9 is controlled by Bessel-renormalized hoppings 0, giving dynamic localization when 1, dynamic delocalization by photon-assisted tunneling, and chiral edge currents generated by second-order next-nearest-neighbor processes with effective plaquette flux
2
In this platform, dynamic localization, dynamic delocalization, and chiral edge currents are read from a single transport spectrum (Xu et al., 2022).
5. Experimental platforms and spectroscopic access
One experimental lineage proceeds from collective matter-wave superradiance in cavities. A Bose–Einstein condensate of 3 placed in an ultra-narrowband optical resonator with 4 and 5 undergoes superradiant scattering into selected discrete momentum states because the cavity linewidth is smaller than the two-recoil energy splitting (Keßler et al., 2014). The pump–cavity resonance condition
6
selects specific links in the discrete momentum graph, and pulse sequences can collectively accelerate or decelerate the whole condensate by multiples of two recoil momenta (Keßler et al., 2014). This system is not yet a superradiance lattice in the tight-binding timed-Dicke sense, but it is a direct bridge between matter-wave superradiance and programmable momentum-site dynamics.
A second lineage uses Bragg-coupled momentum-state lattices of ultracold 7. In the one-dimensional realization, two counter-propagating far-detuned beams couple 8 with spectroscopically resolved control over 9, 0, and 1, and the experiment operates with an open chain of 21 sites read out by 18 ms time-of-flight imaging (An et al., 2017). In higher dimensions, three co-planar beams at 2 generate a triangular momentum lattice 3 with effective Hamiltonian
4
from which honeycomb, square, dice, kagome, Lieb, and decorated-honeycomb momentum-space graphs can be carved out by selective addressing of Bragg resonances (Agrawal et al., 2023). This suggests a direct route to higher-dimensional superradiance-lattice geometries once the external Bragg couplings are replaced or supplemented by collective radiative couplings.
Spectroscopic access has also become highly structured. Injection spectroscopy in a momentum-state lattice introduces a weakly coupled probe site 5 with controllable energy 6, so the loss rate from the probe is
7
an exact analogue of local-density-of-states spectroscopy in scanning tunneling microscopy (Paladugu et al., 2023). The method resolves the eigenenergies of two-site and three-site chains, the band of a 26-site uniform chain, and, in simulation, the Hofstadter butterfly and topological edge states of the Aubry–André–Harper–Hofstadter model (Paladugu et al., 2023). In explicitly superradiant settings, the analogous probes are the superradiant reflectivity and diffraction spectra; in the M-type topological flat-band proposal, the reflectivity
8
tracks the density of states and detects the gap closure at the topological transition (Li et al., 2023).
6. Interactions, collective gauge fields, and conceptual boundaries
In momentum-state lattices of interacting bosons, contact interactions in real space map to effectively attractive, finite-ranged interactions in momentum space. In a mean-field description,
9
so the synthetic lattice acquires a local attractive nonlinearity in momentum space (Gadway et al., 2017). In topological zig-zag ladders with synthetic flux, this nonlinearity stabilizes chiral self-bound states: the wave packet remains localized in momentum space while drifting directionally along the ladder. In a momentum-space double well, the same interaction maps to one-axis twisting,
00
providing a direct route to spin squeezing in a momentum-space lattice (Gadway et al., 2017). For superradiance lattices, the plausible implication is that interaction-induced self-trapping, soliton formation, and squeezing should coexist with collective radiative enhancement once many-body interactions are included.
Not every superradiant cavity problem with momentum dependence is a momentum-space superradiance lattice in the narrow sense. In a fermionic chain with cavity-assisted hopping,
01
the superradiant phase produces an effective band
02
and a persistent directed current 03 on an infinite lattice, but there are no explicit off-diagonal 04 couplings, so the model shifts the momentum-space band rather than constructing a discrete momentum-space lattice graph (Zheng et al., 2016). Likewise, in a gauge-invariant Peierls-substituted Bose–Fermi chain coupled to a single cavity mode,
05
the occupations 06 are conserved and the cavity acts as a dynamical momentum shift, leading to momentum-dependent superradiant transitions, multistability, and squeezed cavity states, but not to a discrete lattice of coupled 07-sites (Su et al., 4 Mar 2025). These models are adjacent to the superradiance-lattice program rather than identical to it.
A plausible implication is that recent Floquet engineering of tight-binding Hamiltonians in momentum-space lattices via quantum resonances of a shaken rotor could be transplanted into future superradiant platforms, because it provides analytical control of momentum-space on-site structures and couplings, including Rice–Mele superlattices, momentum Bloch oscillations, and controlled superlattice periodicity without invoking superradiance explicitly (Ronco et al., 27 Apr 2026). In that sense, momentum-space superradiance lattices now sit at the intersection of timed Dicke physics, synthetic dimensions, Floquet engineering, topological band theory, and collective cavity QED.