Mobile Spin Qubits: Teleportation and Logic in a Reconfigurable Silicon Processor

TL;DR

Researchers at QuTech have bridged a major gap in semiconductor quantum computing by demonstrating that electron spin qubits don't need to stay stationary to interact. By shuttling two electron spins towards each other using "conveyor-mode" transport, they performed a 99% fidelity CZ gate and a 87% fidelity quantum teleportation protocol. This effectively transforms a static 1D array into a dynamic, reconfigurable system.

Background: The Connectivity Bottleneck

Scaling quantum computers is as much about connectivity as it is about qubit count. Most solid-state qubits are "stuck" in place, limited to interacting with their immediate neighbors. To move information across a chip, you usually need a long chain of SWAP gates—which is slow and error-prone.

Taking inspiration from trapped ions (the QCCD architecture), this team asked: What if we move the electrons themselves to a central "interaction zone" to do the logic, and then move them back?

Methodology: Logic on the Move

The device is a 6-quantum-dot array in a ${}^{28}$Si/SiGe heterostructure. The core "magic" happens via Conveyor-Mode Shuttling. Instead of hopping electrons discretely, they apply phase-shifted sinusoidal signals to gate electrodes, creating a moving potential "bucket" that gently carries the electron spin.

1. The Shuttling-Based CZ Gate

The two-qubit interaction (Exchange Coupling, $J$) is distance-dependent. By bringing two mobile qubits (Q2 and Q5) toward the center, their wavefunctions begin to overlap, activating the $J$ interaction.

Figure 1: Conceptual architecture showing storage zones and shared interaction conveyor channels.

2. High-Fidelity Performance

The team achieved a CZ gate in just 58 ns. Through Interleaved Randomized Benchmarking (IRB), they measured a gate fidelity of 98.86%.

A fascinating discovery was the "Merged" potential regime. Usually, $J$ increases exponentially as dots get closer. However, by using elongated potentials, the exchange interaction saturated (plateaued). This regime showed improved dephasing times ($T_2^{\ast }$), suggesting a path to gates that are less sensitive to electrical noise.

Figure 2: Exchange coupling (J) and dephasing times vs. displacement. Note the saturation in the merged configuration.

Achievement: Quantum Teleportation

To prove the utility of mobile qubits for non-local tasks, they implemented quantum state teleportation.

1. Entanglement: Q2 and Q5 were shuttled together to create a Bell state and then separated.
2. Measurement: A Bell-state measurement was performed on Q5 and stationary Q6.
3. Teleportation: This transferred the unknown quantum state of Q6 to the distant Q2.

The 86.7% process fidelity is well above the 66.7% classical threshold, proving that mobile spins can maintain their quantum "spookiness" even after being moved across multiple dot sites.

Figure 3: Interleaved Randomized Benchmarking (IRB) results showcasing the 98.86% CZ gate fidelity.

Critical Insight & Future Outlook

Why is this a big deal? Because it solves the "sparse-to-dense" problem. You can have qubits stored in "quiet" sparse regions and bring them into "busy" dense regions only for logic. This reduces heat, cross-talk, and wiring complexity.

Limitations: The current teleportation is conditional (post-selected) because the readout couldn't distinguish all four Bell states in real-time. The next step is "deterministic" teleportation using fast, non-demolition readout and real-time feed-forward control.

Conclusion: Shuttling is no longer just a way to move data; it’s a high-performance tool for universal quantum logic. This work brings solid-state processors one step closer to the flexible, "all-to-all" connectivity seen in atomic systems.