Quantum Systems Steered into Desired States with New Directional Control

A new method for preparing correlated quantum states advances quantum technologies. Mingsheng Tian and colleagues at The Pennsylvania State University present a dissipative protocol using controllable auxiliary atoms to guide dipolar quantum systems towards specific many-body states without prior knowledge of the system’s Hamiltonian. The approach, designed for neutral atoms in dipolar Rydberg arrays, offers a flexible and scalable framework for state preparation, extending beyond ground state stabilisation and applicable to a wide range of programmable quantum platforms. Enhanced ground state preparation via directional control of dipolar quantum systems Fidelity with target ground states improved from 65% to over 95% using this new protocol, a threshold previously unattainable in many-body systems. This leap enables the preparation of strong, correlated quantum states even with complex interactions, circumventing the need for precise Hamiltonian knowledge. Controllable auxiliary atoms employed to engineer nonreciprocal, energy-selective transitions, effectively steering dipolar quantum systems towards desired states; this directional control represents a sharp advancement over adiabatic methods susceptible to heating. The technique extends beyond ground state stabilisation, providing a scalable framework applicable to various programmable quantum platforms and enabling exploration of excited state many-body physics. Preparation of states with differing particle numbers is now possible, and the stability of n=1, n=2, and n=3 fillings was confirmed. Two types of auxiliary atoms with adjustable detunings accelerated preparation, avoiding delays from scanning a single detuning over time. Nonreciprocal, energy-selective transitions engineered to steer dipolar quantum systems toward desired many-body states, achieving this acceleration. The protocol applied to neutral atoms in dipolar Rydberg arrays, demonstrating its applicability to setups with similar capabilities. Consequently, states across the many-body spectrum stabilised without requiring prior knowledge of the Hamiltonian. Current implementations are limited to small system sizes, and extending to larger, more complex systems presents a challenge. Addressing this limitation will necessitate advancements in computational resources and algorithmic optimisation. Dissipative preparation of tailored many-body quantum states Researchers are currently developing methods for preparing and controlling quantum states, particularly in many-body quantum physics. The emergence of strong correlations and entanglement offers a key resource for quantum technologies. Quantum simulators provide a promising route to explore quantum state engineering due to their precisely controlled and tunable interactions and dynamics. However, the adiabatic approach faces challenges in scenarios requiring tailored symmetry breaking, gap closings in finite-time preparation, and competition from heating in Floquet-engineered systems. As an alternative, dissipative state preparation offers a direct route for state preparation without crossing a critical point. This method evolves the system under engineered dissipation, encoding the target state as the stationary state solution of a governing master equation. These schemes rely on designed dissipative channels and are well suited to highly controllable synthetic quantum platforms. Recent theoretical and experimental progress in superconducting circuits has enabled the dissipative preparation of small-scale microwave-photon Mott insulators. Extending these approaches to other platforms may enable the dissipative preparation of scalable and complex many-body states. Rydberg atom arrays offer a promising platform for realising, tuning, and detecting many-body states due to their flexible geometries and strong, controllable interactions. Advances in the precise trapping and manipulation of neutral atoms in programmable arrays have enabled the exploration of correlated quantum phases and nonequilibrium dynamics. Controlled loss channels have recently been engineered, providing a natural setting for studying open quantum many-body physics. Here, we present dissipative protocols to stabilise many-body correlated states in Rydberg arrays, specifically considering two types of auxiliary atoms, referred to as “source” and “sink”, coupled to the system, featuring state-selective driven-dissipative channels and tunable transition frequencies. These enable nonreciprocal particle excitation and de-excitation at chosen transition energies. By controlling the spectra of the “source” and “sink” atoms, generic initial states are directed to target many-body states through engineered dissipation channels. This provides a framework for preparing many-body ground states with different particle fillings in generic interacting systems, without requiring a priori knowledge of the many-body spectrum, and its efficiency is numerically verified in the interacting dipolar XY model. Spectral engineering protocols for preparing excited many-body states are also provided, enabling the exploration of excited state many-body physics. Our protocol considers a Rydberg atom array, where each atom hosts four relevant states: the ground state |g⟩, a low-lying excited state |i⟩ that decays to |g⟩, and two selected Rydberg states |0⟩ and |1⟩, which encode an effective spin-1/2 degree of freedom and have negligible intrinsic dissipation. Dipolar interactions between Rydberg states are used to engineer the Hamiltonian, for example a spin-1/2 XY model, HS = −Pij(Vij S+i S−j + h.c.), where S+i denotes the spin-raising operator and Vij ∝ 1/|ri − rj|3 is the interaction strength between atoms i and j. The system Hamiltonian is expressed in its eigenbasis as HS = Pλ λ|λ⟩⟨λ|, without restricting to a specific model. The raising operator in this basis is S+i = Pλ,λ′ si,λλ′|λ⟩⟨λ′|, with si,λλ′ = ⟨λ| S+i |λ′⟩. To control such a system, it is coupled to auxiliary atoms via a flip-flop dipolar interaction, HSB = Xij Jije−i∆jtσ−j S+i + h.c. Here, σ−j is the spin-lowering operator for auxiliary atom j and ∆j is the detuning of its resonance [|0⟩↔|1⟩] relative to that of the system atoms. This detuning controls the resonance condition and introduces the phase factor e−i∆jt for spin exchange. Recasting this equation in a rotating frame via U = e−i HSt yields HSB = Xij Xω Jijei(ω−∆j’t σ−j S+i(ω) + h.c., with S+i(ω) = Pλ−λ′=ω si,λλ′ |λ⟩⟨λ′| denoting the raising operator at a specific frequency. Oscillatory off-resonant terms average out under the rotating-wave approximation, yielding an effective transition rate Rλ↔λ′ ∝ δ(ω − ∆j). Central to our stabilisation protocol is the realization of nonreciprocal coupling terms, i.e., Rλ→λ′ ≠ Rλ′→λ, which enables directional preparation of the target state through nonreciprocal, energy-selective state transitions. To this aim, two types of state-selective dissipation for the auxiliary atoms are introduced. One class of “source” atoms are engineered to have rapid decay from |0⟩, while another class of “sink” atoms have large decay from |1⟩. This is realised by resonantly coupling selected Rydberg states to a short-lived intermediate state |i⟩. Owing to rapid spontaneous emission from |i⟩, this coupling induces a tunable effective decay rate (∼Ω20/Γ) from the Rydberg state back to the ground state. In the presence of such dissipation, the spin-exchange processes become effectively nonreciprocal. Source atoms with fast decay from |0⟩ are more likely to undergo exchange with the system atoms by giving up an excitation, with the reverse process being strongly suppressed. Once the source atom reaches the state |0⟩ it rapidly decays and cannot efficiently absorb an excitation from the system. Such a nonreciprocal flip-flop interaction can therefore excite a single particle with a specific frequency in the system. Importantly, optical driving of the source atoms (Ω|g⟩⟨1| + h.c.) replenishes the population of |1⟩, allowing this process to repeat and allowing these auxiliary atoms to act as persistent “sources” of excitations. Conversely, the auxiliary atoms experiencing induced decay from |1⟩ can be dissipatively prepared with high probability in the state |0⟩, allowing them to serve as “sinks” that can efficiently absorb excitations from the system atoms. With the final ingredient of optical AC Stark shifting, to shift the |0⟩↔|1⟩ resonance frequency condition of auxiliary atoms in a locally controlled way, the source and sink atoms can act as energy-selective driven and dissipative channels, respectively. By control of the detuning values (∆j), one can selectively inject or remove excitations at specific Bohr frequencies of the system, enabling energy-resolved nonreciprocal transitions in the system Hilbert space. To stabilise ground states, the spectral response of the auxiliary “source” and “sink” atoms is separately engineered through their detunings ∆j. In the prototypical case of preparing ground states, the source detunings are arranged to cover lower transition energies, while the sink atoms have coverage at higher energies. As a result, there exists a critical energy ωc separating two regimes. For transition energies below ωc, the system is more likely to absorb “source” excitations, being directed from eigenstates with occupancy n to n + 1. Conversely, above ωc, the “sink” coupling dominates and drives system transitions with ∆n = −1. To characterise its stabilisation mechanism, work is done in the rotating frame defined by U = eiωct P i ni, where an eigenstate in the n-particle sector is considered. Computational limits hinder scaling of quantum system simulations Precise state preparation demands control of complex quantum systems, yet achieving this reliably across the many-body spectrum presents a persistent challenge. This new protocol offers a route to stabilise states without prior knowledge of the system itself, but simulations currently rely on the quantum trajectory method, a computationally intensive technique. The scaling of the necessary Hilbert space quickly becomes problematic, demanding significant resources even for modest system sizes. This new protocol establishes a fundamentally new approach to controlling quantum states, offering a pathway to stabilisation without needing detailed prior knowledge of the system itself. Controllable ‘auxiliary atoms’ steer quantum systems towards desired configurations, effectively creating a directional path within the system’s complex possibilities. This work establishes a new method for preparing many-body quantum states, circumventing the need for detailed prior knowledge of a system’s Hamiltonian. By employing these controllable ‘source’ and ‘sink’ atoms, the protocol directed quantum systems towards target states via nonreciprocal, energy-selective transitions, effectively creating a directional pathway within the system’s possibilities. This protocol, demonstrated in simulations of dipolar Rydberg atom arrays, offers a scalable framework extending beyond ground state stabilisation and applicable to diverse programmable quantum platforms. The researchers demonstrated a new protocol for preparing correlated quantum states in many-body systems without requiring prior knowledge of the system’s Hamiltonian. This method utilises controllable ‘auxiliary atoms’ to steer quantum systems towards desired states via energy-selective transitions, effectively creating a directional pathway within the system’s possibilities. The protocol was demonstrated through simulations using dipolar Rydberg atom arrays and offers a scalable framework for state preparation. The authors note that current simulations are limited by computational demands as the Hilbert space scales with system size. 👉 More information🗞 Dissipative Preparation of Correlated Quantum States in Dipolar Rydberg Arrays🧠 ArXiv: https://arxiv.org/abs/2604.18542 Tags: