How patterned quantum lattices could sharpen sensing

A sensor detects an unknown quantity by measuring how it changes a system. Quantum sensors use quantum states to make that change easier to distinguish. In this theoretical study, Saubhik Sarkar, Chiranjib Mukhopadhyay and Abolfazl Bayat ask whether a patterned lattice can improve that process. They focus on non-Hermitian systems, which can describe open systems and, in the models considered here, include unequal couplings for motion to the left and right.

The researchers study one-dimensional chains made from repeating modules. Each module contains sites connected by one set of couplings, while another set links neighboring modules. Some couplings take distinct values at regular positions in the repeating structure. The authors analyze models related to the Hatano–Nelson and Su–Schrieffer–Heeger chains using analytical calculations and numerical simulations.

In these models, unequal left- and right-directed couplings introduce non-reciprocity: the two directions do not have equivalent coupling strengths. This can produce the non-Hermitian skin effect, in which many modes become concentrated near an edge when the chain has open ends. The authors show that the modular couplings change the conditions for this effect and for related topological transitions. In particular, they derive a condition for a point gap in the complex-valued energy spectrum to close. Varying the modular couplings can shift where that transition occurs.

The unusual behavior of modes also complicates the standard connection between a system’s bulk properties and its edge states. To study band topology, the paper uses a generalized Brillouin-zone method, designed to account for these non-Hermitian effects. The calculations indicate that modular structure can introduce additional possibilities for band gaps to close and alter the winding-number description of phases. But not every gap closing is useful for sensing: the authors find that line-gap transitions do not generally produce the localization-to-delocalization change their sensing strategy relies on.

For sensing, the team calculates how much information quantum states carry about parameters in the couplings. Their main probe is a steady state associated with the eigenstate whose energy has the largest imaginary part. They also examine states produced by evolving from a localized starting state for a finite time. Measures such as quantum Fisher information describe an idealized limit on estimation precision. They do not, on their own, show that an experimental sensor achieves that precision.

Near a point-gap transition, the relevant state changes from edge-localized to delocalized. The authors find that this change can make the state more responsive to the parameter driving the transition. In selected models and parameter ranges, modular versions show higher calculated sensitivity than the corresponding non-modular models. For some single-parameter cases, the reported information grows approximately quadratically with system size. The study also analyzes simultaneous estimation of two parameters and a three-parameter scenario enabled by modular couplings. These gains are conditional: outside favorable ranges, performance can be worse than the non-modular comparison, even where enhanced scaling remains.

There are practical qualifications. As the system grows, the sensitivity peak becomes narrower, making it important to locate the transition accurately. Preparing the desired state takes time. In a proposed photonic quantum-walk implementation, non-unitary evolution would involve losses and post-selection. The authors estimate aspects of feasibility but do not report a new experiment; they also identify mapping the modular couplings onto experimental controls as a task for future work.

The study therefore suggests a design strategy, not a universal promise: selected couplings in a repeating pattern can provide control over transition locations and state responses near them. Its conclusions rely on specific models, finite-size numerical analyses and idealized precision measures. Experimental results could depend on state preparation, measurement, losses and implementation details. This article describes an arXiv preprint, which may not have undergone peer review; the available evidence does not establish independent validation of these particular findings.