07/16/2026
By Razvan Stanescu

The Kennedy College of Sciences, Department of Physics, invites you to attend a doctoral dissertation defense by Razvan Teodor Stanescu on "Effective Hamiltonians for Quantum Control of Driven Transmon Qubits: Pulse Correction, Leakage, and Crosstalk."

Candidate Name: Razvan Teodor Stanescu
Degree: Doctoral
Defense Date: Monday, Aug. 3, 2026
Time: Noon to 1:30 a.m.
Location: Room 110, Lydon Library, North Campus, UMass Lowell and via Zoom

Thesis/Dissertation Title: Effective Hamiltonians for Quantum Control of Driven Transmon Qubits: Pulse Correction, Leakage, and Crosstalk

Committee Members:

  • Advisor: Hugo Ribeiro, Ph.D., Department of Physics and Applied Physics, University of Massachusetts Lowell
  • Marian Jandel, Ph.D., Department of Physics and Applied Physics, University of Massachusetts Lowell
  • Archana Kamal, Ph.D., Department of Physics and Astronomy, Northwestern University

Brief Abstract:

Reliable quantum control depends not only on the design of control protocols, but also on the physical models used to derive them. This raises a fundamental question: How does the choice of model Hamiltonian affect our ability to predict and control quantum dynamics? Physically well-justified approximations may reproduce broad features of a system while failing to capture dynamical quantities that are critical for control. We investigate this issue for superconducting transmon qubits by comparing the commonly used Duffing-oscillator approximation with a more accurate representation constructed in the eigenbasis of the transmon Hamiltonian. We show that discrepancies between these descriptions can lead to substantially different predictions of the coherent dynamics and, consequently, of the control fields required to implement high-fidelity operations.

A second challenge concerns the complexity of analytically derived control pulses. Even for a comparatively simple system such as a transmon, perturbative control methods can produce pulses containing many Fourier components. This complexity reflects the large number of coherent error channels that must be addressed, including leakage outside the computational subspace and phase errors within it. We therefore ask: Can one construct simple control pulses with minimal, experimentally natural spectral content without resorting to direct fidelity optimization or imposing constraints through a black-box procedure?

To address this question, we develop a physically transparent strategy for identifying and correcting the dominant error channels. The method uses a perturbative description of the dynamics to connect individual pulse components to specific physical errors, while searching for corrections within a deliberately restricted and experimentally meaningful control space. In this way, pulse simplicity is built into the construction rather than enforced through constrained fidelity optimization. Applied to transmon gate control, the resulting approach produces compact control pulses while retaining a clear account of the origin and cancellation of coherent errors. More broadly, this work demonstrates that accurate modeling and physically interpretable pulse design are inseparable components of reliable quantum control.