Lawrence Livermore National Laboratory



Using Quantum Probes to Understand the Material Origins of Decoherence


Materials-induced decoherence is generally considered a necessary evil of quantum computing. An LLNL team is working to change that.

The impact of material-induced noise

To move quantum computing from theory to application, researchers have developed a diverse set of solid-state quantum platforms—most notably superconducting qubits, quantum dots, and crystal defects.

Unfortunately, all of these solid-state systems are subject to decoherence due to noise emanating from surfaces and the bulk. The result is degraded quantum sensor performance and limited quantum gate fidelity.

Until now, researchers have primarily treated this materials-induced decoherence as an unfortunate fact of nature. As a result, qubits (quantum bits) have been engineered to be as insensitive as possible to their surrounding environments.

While this approach can increase the coherence times of qubits and quantum sensors, it negatively impacts the sensors' usefulness by weakening their coupling to sensing targets.

Likewise, restricting qubits' environmental sensitivity impedes the scaling up of quantum computation devices by reducing the strength of qubit-qubit coupling. That limits the parameter space over which qubits can be tuned and can increase the physical footprint of the qubits themselves.

Using quantum probes to reduce decoherence

Our project leverages quantum probes to determine the microscopic origins of decoherence. The data we gather will improve our control over materials-induced noise, enabling an order-of-magnitude increase in coherence times.

With enhanced qubit design and error correction protocols in place, we should realize significant improvements in quantum sensor sensitivity and spatial resolution as well as reduced physical qubit overhead, opening the door to new advances in fault-tolerant quantum computation.

People

  • Group members that are part of the project: Yaniv Rosen, Luis Martinez, Daniel Tennant, Sean O'Kelley, Jonathan DuBois
  • Internal collaborators: Keith Ray, Kevin Chaves, Artur Tamm, Vince Lordi
  • External collaborators: Victor Brar, Jennifer Choy, Mark Eriksson, Lara Faoro, Mark Friesen, Shimon Kolkowitz, Alex Levchenko, Robert McDermott
    • Affiliation: University of Wisconsin - Madison
  • Funding info: 09/2020 – 09/2023

Three graphs showing how quantum state data differs when 1) recorded raw, 2) analyzed using conventional classifiers, 3) analyzed with a Hidden Markov Model classifier.
Quantum state readout with Hidden Markov Models (HMM).



Two graphs showing the change in charge states over time.
Identifying charge states with Hidden Markov Model (HMM).



A series of four images illustrating the types of information made available by LLNL's quantum probe.
Figure a: Electron affinity calculations of oxygen on a diamond surface for NV center qubits. Figure b: Oxygen molecules on an aluminum oxide surface to study flux noise. Figure c: Charge state transition levels for Si and Ge dangling bonds for quantum dot qubits. Figure d: Josephson junction model.