Dominique Sugny

Quantum optimal control has become a central tool for designing high-performance pulse sequences in magnetic resonance and quantum technologies. Early developments in NMR relied on heuristic composite and shaped pulses or analytical approaches such as average Hamiltonian theory. The introduction of modern optimal control algorithms enabled the systematic optimization of thousands of control parameters, greatly expanding the achievable performance and scope of pulse engineering. This review summarizes key developments in optimal control of spin and pseudo-spin systems and highlights major applications in magnetic resonance, in quantum technologies, and in the control of atoms and molecules.

Narrowband optimal control is proposed as a protocol of optimal quantum sensing, providing maximum accuracy for given resources. The procedure is demonstrated on the IBM quantum computer to estimate the amplitude of the qubit Rabi frequency. It is based on optimizing the quantum Fisher information (QFI) when it becomes equal to the classical Fisher information. This scenario offers a practical quantum sensing because it only involves population transfer to the excited state as the target without dynamical phases, which would otherwise be difficult to control and measure. The protocol uses the single-shot-shaped pulse method and an inverse optimization, to design a pulse that results in the optimal QFI in terms of energy or equivalently operating time, referred to in short as local narrowband inverse optimization (NIO). Multiple π pulses are shown to be particular cases of this protocol. An alternative semiglobal optimal NIO procedure producing a partially nonoscillatory profile is also demonstrated. Our approach also offers application to optimal atomic selective excitation in an array of atoms for quantum technology.

Protecting qubits from environmental noise while maintaining strong coupling for fast high-fidelity control is a central challenge for quantum information processing. Here, we demonstrate a control scheme for superconducting fluxonium qubits that eliminates qubit decay through the control channel by suppressing the environmental density of states at the transition frequency. Adding a low-pass filter on the flux line allows for flux-biasing and, at the same time, coherently controlling the fluxonium qubit by parametrically driving it at integer fractions of its transition frequency. We compare the filtered to the unfiltered configuration and find a 5-times-longer T 1 , and a 10-times-improved T 2 -echo time in the filtered case. We demonstrate coherent control with up to 11-photon subharmonic drives, highlighting the strong nonlinearity of the fluxonium potential. Measured Rabi frequencies and drive-induced frequency shifts show excellent agreement with numerical and analytical models. Furthermore, we show the equivalence of a 3-photon subharmonic drive to an on-resonance drive by benchmarking subharmonic gate fidelities above 99.94%. These results open up a scalable path for full qubit control through a single Purcell-protected channel, providing strong suppression of control-induced decoherence and enabling wiring-efficient superconducting quantum processors.

In this paper, we design and experimentally implement various robust quantum unitary transformations (gates) acting on d -dimensional vectors (qudits) by tuning a single control parameter using optimal control theory. The quantum state is represented by the momentum components of a Bose-Einstein condensate placed in an optical lattice, with the lattice position varying over a fixed duration serving as the control parameter. To evaluate the quality of these transformations, we employ standard quantum process tomography. In addition, we show how controlled unitary transformations can be used to extend state stabilization to global stabilization within a controlled vector subspace. Finally, we apply them to state tomography, showing how the information about the relative phase between distant momentum components can be extracted by inducing an interference process.

We study different characterizations of P-divisibility in two-level open quantum systems whose dynamics are governed by a time-local master equation with time-dependent relaxation rates. Necessary and sufficient conditions for the P-divisibility of the dynamical map are given in terms of inequalities on such relaxation rates. The equivalence between several P-divisibility characterizations existing in the literature is explicitly proven. The connection to the Breuer-Laine-Piilo measure of non-Markovianity is also established. As an application of such characterizations, we study the open dynamics of a qubit interacting with a bosonic mode. More precisely, we characterize the properties of the local map on the qubit generated by its interaction with the bosonic mode, playing the role of an extremely reduced bath. Interesting observations are made, opening perspectives for a deeper physical understanding of CP-divisibility, P-divisibility, and BLP measure.

We study different characterizations of P-divisibility in two-level open quantum systems whose dynamics are governed by a time-local master equation with time-dependent relaxation rates. Necessary and sufficient conditions for the P-divisibility of the dynamical map are given in terms of inequalities on such relaxation rates. The equivalence between several P-divisibility characterizations existing in the literature is explicitly proven. The connection to the Breuer-Laine-Piilo measure of non-Markovianity is also established. As an application of such characterizations, we study the open dynamics of a qubit interacting with a bosonic mode. More precisely, we characterize the properties of the local map on the qubit generated by its interaction with the bosonic mode, playing the role of an extremely reduced bath. Interesting observations are made, opening perspectives for a deeper physical understanding of CP-divisibility, P-divisibility, and BLP measure.