Dissipative 1D Hubbard Model Insights

Updated 23 February 2026

The paper demonstrates how introducing Lindblad-type dissipation into the 1D Hubbard framework enables the study of non-equilibrium steady states and quantum phase transitions.
It employs advanced numerical and analytical methods, including MPO, exact diagonalization, and Bethe ansatz, to quantify relaxation dynamics and identify critical dissipative gaps.
Engineered dissipation is shown to facilitate precise quantum state preparation and controlled transport, bridging theoretical models with experimental ultracold atom implementations.

The dissipative one-dimensional (1D) Hubbard model is a framework for studying quantum many-body systems in the presence of engineered or natural dissipation. Incorporating dissipation channels into the Hubbard model—a paradigmatic lattice Hamiltonian for strongly correlated particles—enables exploration of non-equilibrium steady states, real-time relaxation processes, decoherence, and the active engineering of quantum phases and entanglement. The dissipative 1D Hubbard model encompasses both bosonic and fermionic realizations, described by Lindblad master equations where coherent evolution is modified by jump operators reflecting loss, dephasing, inelastic collisions, or coupling to quantum baths. This setting connects quantum optics, many-body theory, and quantum simulation with ultracold atoms, providing both analytic and experimental access to non-equilibrium quantum phenomena inaccessible in purely closed systems.

1. Defining Hamiltonians and Lindblad Dynamics

The 1D dissipative Hubbard model generalizes the canonical closed-system Hubbard model by incorporating Lindblad-type dissipative processes. The Hamiltonian for bosons (Bose-Hubbard) or fermions (Fermi-Hubbard) on a 1D chain of $M$ sites is

$H = -J\sum_{\langle i,j\rangle,\sigma} (c_{i\sigma}^\dagger c_{j\sigma} + \mathrm{h.c.}) + U\sum_{j} n_{j\uparrow} n_{j\downarrow} + \sum_j \varepsilon_j n_j$

with $c_{i\sigma}$ annihilating a boson or fermion of spin $\sigma$ on site $i$ , $J$ the tunneling amplitude, $U$ the on-site interaction, and $\varepsilon_j$ the site energy. The system's open quantum dynamics obeys a Gorini–Kossakowski–Sudarshan–Lindblad master equation,

$\frac{d\rho}{dt} = -i[H,\rho] + \sum_\ell \gamma_\ell \Big( L_\ell \rho L_\ell^\dagger - \frac{1}{2} \{L_\ell^\dagger L_\ell, \rho\}\Big)$

where the $L_\ell$ specify dissipative channels: single-particle loss ( $L_j = c_j$ ), phase noise ( $L_j = n_j$ ), two-body loss ( $L_j = c_{j\downarrow} c_{j\uparrow}$ ), and more (Kordas et al., 2015, Kordas et al., 2015, Nakagawa et al., 2020, Sponselee et al., 2018).

Nonlocal and engineered jump operators enable targeted dissipative stabilization—e.g., bond dissipators $L_j = (b_j^\dagger + b_{j+1}^\dagger)(b_j - b_{j+1})$ for condensate generation (Bonnes et al., 2014), or bidirectional Raman-induced hopping for antiferromagnetic order (Kaczmarczyk et al., 2016). Coupling each site to Markovian or non-Markovian baths broadens the phenomenology further (Ribeiro et al., 2023, Bouverot-Dupuis et al., 12 Nov 2025).

2. Physical Phenomena: Decoherence, Dark States, and Steady-State Engineering

Dissipation in 1D Hubbard models yields a rich set of dynamical and steady-state phenomena:

Quantum Zeno effect: Strong local loss or measurement suppresses tunneling into lossy sites, stabilizing insulating states and producing non-monotonic residual population as a function of the loss rate (Kordas et al., 2015, Kordas et al., 2015).
Breather and soliton formation: For bosons, strong local loss spatially confines particles, creating long-lived discrete breathers or dark solitons as metastable or steady states (Kordas et al., 2015, Ceulemans et al., 2023).
Dissipation-induced entanglement and many-body dark states: Engineered dissipation can prepare highly entangled Schrödinger-cat–like superpositions, such as Dicke states that are “dark” to further two-body loss in Fermi-Hubbard chains (Sponselee et al., 2018). Specific jump operator structures allow for the dissipative preparation of pure antiferromagnetic Néel states as unique dark steady states, even in 1D, through bidirectional and unidirectional density-dependent hopping channels (Kaczmarczyk et al., 2016).
Destruction or stabilization of long-range order: Generic Markovian or Ohmic dissipation destroys equilibrium Luttinger-liquid and Mott-insulator phases, often driving transitions to new steady states characterized by long-range order or infinitely compressible but non-superfluid “Mott*” states (Ribeiro et al., 2023, Bouverot-Dupuis et al., 12 Nov 2025). Conversely, tailored dissipation can robustly stabilize ordered phases.
Transport and non-equilibrium currents: Dissipation enables true steady-state currents under applied electric fields or chemical “voltages,” suppressing Bloch oscillations and allowing detailed study of non-equilibrium transport and blockades (Neumayer et al., 2015, Kordas et al., 2015).

3. Analytical and Numerical Solution Techniques

The study of dissipative 1D Hubbard models deploys a range of numerical and analytical frameworks:

Exact diagonalization and quantum-jump Monte Carlo: Tractable for small system sizes, these methods reveal quantum trajectories, relaxation rates, entanglement growth, and allow extraction of waiting-time distributions (Kordas et al., 2015, Kaczmarczyk et al., 2016, Schaller et al., 2021, Ceulemans et al., 2023).
Matrix product operator (MPO) and matrix product state (MPS) techniques: Efficient for dissipative bosonic chains, these access both dynamical properties and steady states, and can probe the scaling of operator-space entanglement entropy (Bonnes et al., 2014).
Mean-field, Gutzwiller, and truncated Wigner approaches: These approaches approximate the quantum dynamics in the large-occupation regime, capturing bistability, phase transitions, and quantum noise-induced switching (Kordas et al., 2015, Ceulemans et al., 2023).
Functional renormalization group (FRG): Provides a nonperturbative, systematically improvable framework for the full dissipative phase diagram in the presence of non-Markovian, power-law spectral baths, yielding universal critical exponents and identifying transitions between Luttinger-liquid, dissipative fixed points, and long-range ordered phases (Bouverot-Dupuis et al., 12 Nov 2025).
Exact solutions via non-Hermitian Bethe ansatz: For Hubbard models with two-body loss or dephasing, the Liouvillian can be diagonalized, yielding full spectral information, steady states, Liouvillian gaps, exceptional points (with diverging correlation lengths), and quantum Zeno-induced spin-charge separation (Medvedyeva et al., 2016, Nakagawa et al., 2020). For the Fermi-Hubbard model, effective non-Hermitian Hamiltonians allow direct calculation of relaxation rates, time scales, and decay of order parameters in finite clusters and beyond (Schaller et al., 2021).
Quantum Monte Carlo with wormhole updates: For non-Markovian, retarded interactions induced by local oscillator baths, wormhole QMC yields unbiased results for phase boundaries and critical phenomena (Ribeiro et al., 2023).

4. Dissipative Phase Diagrams: Order, Instabilities, and Phase Transitions

Dissipative 1D Hubbard models exhibit rich phase diagrams controlled by interaction strength, bath coupling, and dissipation parameters:

Ohmic and super-Ohmic baths: In Bose-Hubbard chains, coupling to Ohmic baths destabilizes the Luttinger-liquid at infinitesimal coupling, producing a long-range superfluid phase; Mott insulators remain distinct only at small bath coupling and become infinitely compressible, before ultimately transiting to superfluid order. Super-Ohmic baths preserve Luttinger-liquid behavior up to a finite critical coupling, beyond which order emerges (Ribeiro et al., 2023, Bouverot-Dupuis et al., 12 Nov 2025).
Berezinskii–Kosterlitz–Thouless transitions: FRG analyses demonstrate that the competition between Luttinger-liquid and dissipative fixed points is separated by a BKT transition at a critical Luttinger parameter $K_c$ that depends on bath exponent $s$ (Bouverot-Dupuis et al., 12 Nov 2025).
Non-Hermitian phase transitions and exceptional points: In the presence of loss, the spectrum of non-Hermitian effective Hamiltonians can exhibit $\mathcal{PT}$ -symmetry breaking, exceptional points with merged eigenvalues, and associated divergences in correlation lengths (Pan et al., 2020, Nakagawa et al., 2020). The Mott regime retains real spectra (no $\mathcal{PT}$ breaking) in its low-energy sector, while higher bands can display dynamical instabilities.
Long-range order from dissipation: Engineered local or nonlocal dissipation can realize pure steady states—antiferromagnetic or condensate states—as unique dark states, even where equilibrium statistical mechanics would prohibit such ordering (e.g., Mermin-Wagner violation in 1D via driven-dissipative processes) (Kaczmarczyk et al., 2016, Bonnes et al., 2014).

5. Relaxation Dynamics, Time Scales, and Critical Gaps

Dissipative 1D Hubbard models display multi-scale relaxation:

Separation of time scales: In Fermi-Hubbard systems with bath coupling, charge relaxation occurs rapidly ( $\tau_\mathrm{charge}\sim1/\gamma$ ), enabling rapid stabilization of Mott-insulating density, while spin order decays on a much slower scale set by virtual processes ( $\tau_\mathrm{spin}\sim (U/J)^2 (1/\gamma)$ ), resulting in an extended persistence of Mott physics without magnetic order (Schaller et al., 2021).
Liouvillian spectrum and gaps: The slowest nonzero Liouvillian eigenvalue, defining the dissipative gap, governs the asymptotic approach to steady state. In many scenarios, especially under global or bond dissipation, this gap closes algebraically as $L^{-2}$ in the thermodynamic limit, signaling critical slowing down and the absence of a true spectral gap (Bonnes et al., 2014, Ceulemans et al., 2023, Medvedyeva et al., 2016).
Dynamical instability and $\mathcal{PT}$ -symmetry breaking: Instabilities can occur as dissipation or interactions are tuned, generating exponential growth in certain density matrix components—interpreted as $\mathcal{PT}$ symmetry breaking in the effective non-Hermitian Hamiltonians (Pan et al., 2020).
Exceptional points and diverging correlation lengths: Analytic continuation of Bethe-ansatz solutions yields exact expressions for the correlation length, which can diverge at special (exceptional) parameter points where the spectrum becomes nondiagonalizable (Nakagawa et al., 2020).

6. Quantum State Engineering and Transport Control

The dissipative 1D Hubbard model provides a route for engineered quantum state preparation and controlled transport:

Dissipative preparation of antiferromagnetic order: By combining bidirectional Raman-induced hopping and unidirectional “filling” channels, one can deterministically drive a 1D Fermi-Hubbard chain into a pure Néel state, with high steady-state order parameters ( $I_\mathrm{AF} \gtrsim 0.8$ ) controlled by the ratio of dissipative to coherent rates (Kaczmarczyk et al., 2016).
Dark Dicke states via strong two-body loss: In Fermi-Hubbard chains with two-body dissipation, initial two-body loss is quenched dynamically by the emergence of strongly entangled Dicke states (fully symmetric in spin, antisymmetric in spatial wavefunction), halting further decay—an experimentally confirmed mechanism (Sponselee et al., 2018).
Non-equilibrium transport and resonances: Fermionic 1D Hubbard models with site-resolved baths support steady currents under applied fields, with current resonances arising from underlying antiferromagnetic correlations and the formation of Wannier–Stark ladders. Dissipation both suppresses unwanted coherent oscillations and broadens resonance features, tunable via bath coupling (Neumayer et al., 2015).
Preparation and control of bosonic dark and breather states: In the Bose-Hubbard context, engineered dissipation yields pure condensates, breathers, or dark solitons, with dissipation fixing coherence properties and the refilling or draining of selected sites (Kordas et al., 2015, Ceulemans et al., 2023).

7. Experimental Realizations and Observables

Experimental implementation of the dissipative 1D Hubbard model exploits tools from ultracold atomic physics:

Engineering jump operators: Laser-induced Raman transitions, optical pumping, and controlled inelastic collisions enable the realization of desired dissipative processes, including density-dependent hopping, unidirectional refilling, and two-body loss (Kaczmarczyk et al., 2016, Sponselee et al., 2018, Kordas et al., 2015).
Observation of dynamical phenomena: Atom loss curves, site-resolved density correlations, modulation spectroscopy, and quantum gas microscopy provide direct probes of relaxation rates, dark-state preparation, steady-state currents, and the onset of order or decoherence (Sponselee et al., 2018, Schaller et al., 2021).
Tuning system parameters: Optical lattice depths (control over $J/U$ ), Feshbach resonances (control over $U$ ), and bath engineering via external fields or tailored environments allow comprehensive access to the parameter regimes described in theoretical work.
Scaling and time scale observables: The algebraic closing of dissipative gaps and slowing down near phase transitions are accessible via monitoring relaxation of order parameters and fluctuations as system size increases (Bonnes et al., 2014, Ceulemans et al., 2023).