Reducing Decoherence, Alternating Bias Assisted Annealing Improves Amorphous Oxide Tunnel Junctions

Amorphous oxide materials form the foundation of many superconducting qubits, but imperfections within these materials can limit qubit performance by causing energy loss. Alexander C. Tyner from University of Connecticut and Stockholm University, along with Alexander V. Balatsky from University of Connecticut, investigate a technique called alternating bias assisted annealing (ABAA) to address this challenge. Their research replicates an accelerated ABAA process using advanced computer simulations, combining fundamental physics with machine learning to understand how applying alternating electrical signals alters the energy landscape of these materials. This work reveals the mechanisms by which ABAA reduces defects, paving the way for more stable and reliable superconducting qubits and advancing the field of quantum computing.

Conducting quantum bits (qubits) presents significant challenges, particularly those associated with defects and sample variance among the tunneling barriers. The methodology of alternating bias assisted annealing (ABAA) addresses these issues, demonstrably reducing defects that give rise to two-level systems which couple to the qubit and expedite decoherence, limiting performance. This work replicates an expedited ABAA process through a combination of ab-initio molecular dynamics and machine-learned potentials, illuminating how ABAA affects the energy landscape of the barrier.

Amorphous Oxide Junctions and Decoherence Origins

This research investigates the origin of decoherence in superconducting qubits and explores a method to mitigate it. Superconducting qubits, promising candidates for quantum computing, suffer from decoherence, the loss of quantum information. A major source of this decoherence is believed to be two-level systems (TLS) within the amorphous oxide tunnel junctions used in qubits, acting as parasitic quantum systems that interact with the qubit, causing it to lose coherence. The exact nature and origin of these TLS are still debated, but dangling bonds and structural defects in the amorphous oxide are suspected.

Researchers explored ABAA, where a time-varying electric field is applied to the amorphous oxide tunnel junction. ABAA was found to reduce the density of TLS, thereby improving qubit coherence. The alternating electric field drives transitions between the energy levels of the TLS, causing the TLS to rearrange, effectively healing or passivating some of the defects responsible for the TLS. Researchers used advanced computational methods to simulate the structure and properties of the amorphous aluminum oxide used in the tunnel junctions and investigate how the electric field affects the TLS and the underlying defect structure, supporting the proposed mechanism of defect rearrangement and TLS passivation.

ABAA is a promising technique for improving qubit coherence by reducing the density of TLS in amorphous oxide tunnel junctions. First-principles simulations provide valuable insights into the underlying physics and support the proposed mechanism. This research contributes to a better understanding of decoherence in superconducting qubits and paves the way for developing more robust and reliable quantum computing devices.

Amorphous Oxide Barriers and Defect Reduction Mechanisms

Scientists have achieved a detailed understanding of the atomic processes occurring within amorphous oxide tunneling barriers, crucial components in superconducting qubits. This work replicates an expedited alternating bias assisted (ABAA) process, a technique used to reduce defects that cause decoherence in qubits, by combining ab-initio molecular dynamics with machine-learned potentials. The research illuminates how ABAA affects the energy landscape of the barrier, revealing the mechanisms behind its effectiveness. The team generated amorphous aluminum oxide structures using a melt and anneal protocol, beginning with a crystalline sample.

Atoms were randomly displaced and heated to create a liquid-like state, then quenched and annealed to create a stable amorphous structure suitable for simulating tunneling barriers. To model the application of an alternating bias, scientists utilized Car-Parinello molecular dynamics. Simulations tracked the total energy of the amorphous oxide barrier over time, revealing that a positive bias plateaus after approximately 2 picoseconds. Subsequent simulations demonstrated that allowing the system to age without bias resulted in a stable energy state, while applying a bias of opposite sign and equal magnitude also led to stabilization.

Measurements of the total energy during the simulation revealed that the system reaches a plateau after approximately 2 picoseconds when a positive bias is applied. This research provides insight into the necessary timescales for the ABAA procedure and how it allows traversal of the energetic landscape of the tunneling barrier, identifying new energetic minima consistent with experimental findings. The computational methodology achieves timescales of 1-10 picoseconds, despite the challenges of simulating ABAA protocols that can last for many seconds in experiments.

ABAA Stabilises Oxide Barriers, Reduces Defects

This work investigates the impact of alternating bias assisted annealing (ABAA) on amorphous oxide tunneling barriers, a crucial component in superconducting qubits. Researchers successfully replicated the effects of ABAA using a combination of advanced computational techniques, including ab-initio molecular dynamics and machine-learned potentials. Their simulations illuminate how the ABAA process alters the energy landscape of the oxide barrier, specifically reducing the density of localized vibrational modes, thought to be the source of performance-limiting defects known as two-level systems. The findings demonstrate that ABAA effectively moves the oxide barrier towards a more stable, lower-energy state, diminishing the number of problematic defects that can disrupt qubit coherence.

Importantly, the study suggests that ABAA does not eliminate these defects entirely, but rather shifts their frequency outside the range where they can significantly interact with the qubit. This represents a refinement of the barrier’s properties, improving qubit performance by minimizing disruptive influences. Acknowledging the computational demands, the authors highlight future research directions focused on optimizing the ABAA protocol through variations in temperature, bias magnitude, and the number of applied alternating biases. Further investigation along these lines promises to enhance the effectiveness of this technique for improving the stability and coherence of superconducting qubits, advancing the field of quantum computing.