Proposal for a Loophole-Free Bell Test with Electron Spins of Donors in Silicon

Fang-Yu Hong, Shi‐Jie Xiong, Yang Xiang, Weihua TangJournal of Modern Physics 2011 · DOI

Abstract: So far, all experimental tests of Bell inequalities which must be satisfied by all local realistic hidden-variable theories and are violated by quantum mechanical predictions have left at least one loophole open. We propose a feasible setup allowing for a loophole-free test of the Bell inequalities. Two electron spin qubits of donors $^{31}P$ in a nanoscale silicon host in different cavities 300 m apart are entangled through a bright coherent light and postselections using homodyne measurements. The electron spins are then read out randomly and independently by Alice and Bob, respectively, with unity efficiency in less than 0.7 µs by using optically induced spin to charge transduction detected by radio-frequency single electron transistor. A violation of Bell inequality larger than 37% and 18% is achievable provided that the detection accuracy is 0.99 and 0.95, respectively.

Introduction: Closing the Loopholes

Since John Bell first proposed his famous inequalities in 1964, physicists have sought to prove that quantum mechanics is fundamentally "non-local." This means particles can influence each other instantaneously across space in ways that "local realism"—the idea that objects have definite properties independent of observation and only interact with their immediate surroundings—cannot explain.

While many experiments have shown violations of Bell’s inequalities, they almost always suffer from two major "loopholes":

  1. The Detection Loophole: If detectors are inefficient, we only see a small sample of events. Skeptics can argue that the "missing" data might have obeyed local realism.
  2. The Locality (Lightcone) Loophole: If the measurement stations are too close together or the measurement takes too long, a signal (traveling at or below the speed of light) could theoretically pass between the stations and coordinate the results.

To truly settle the debate, an experiment must close both loopholes simultaneously. This paper proposes a feasible experimental setup to perform such a "loophole-free" Bell test using electron spins of phosphorus donors in silicon.

Proposed Experimental Architecture

The authors propose using the Kane Si:P architecture, which involves phosphorus-31 ($^{31}P$) donor atoms embedded in a silicon host. This solid-state approach offers a significant advantage: silicon-based donor spins are stable, can be read out with nearly 100% efficiency, and can be integrated into existing semiconductor fabrication workflows.

Schematic of the proposed Si:P device architecture showing the donor atoms, control gates, and readout SET.

In this setup, two electron spin qubits are placed in high-quality (high-Q) cavities separated by a distance of 300 meters. They are entangled using pulses of bright coherent light. The light interacts weakly with the donors, and entanglement is confirmed through homodyne measurements and post-selection. By placing the qubits 300 meters apart, the setup ensures that no signal—even one traveling at the speed of light—could travel between the measurement stations (Alice and Bob) to influence their results during the measurement window.

High-Speed Readout and Efficiency

To close the locality loophole, the readout must be faster than the time it takes light to travel the 300-meter separation distance (~1 $\mu s$). The authors utilize optically induced spin-to-charge transduction to achieve this. This process converts the spin state into a charge state, which is then detected by a radio-frequency single electron transistor (rf-SET).

Numerical simulations indicate that this readout method offers near-unity efficiency, which is critical for closing the detection loophole. The spin-to-charge transduction and rf-SET detection can be completed in approximately 0.6 $\mu s$, comfortably within the 1 $\mu s$ limit required for the 300m separation.

Key Results and Predicted Violations

The model predicts a significant violation of the CHSH (Bell) inequality. Specifically, a violation of 37% is achievable if detection accuracy is 0.99, and 18% if accuracy is 0.95.

Predicted Bell inequality violation (B), fidelity (F), and success probability (Ps) as a function of the parameter Pc.

With a repetition rate of 500 kHz, the authors estimate that a complete, statistically significant Bell test could be finished in less than one second. This speed and robustness to detection inaccuracy (maintaining a violation even at 95% accuracy) make the Si:P platform a highly promising candidate for definitive non-locality testing.

Challenges and Future Directions

While the paper argues the setup is within the reach of current technology, several engineering challenges remain:

  • Fabrication Precision: The proposal relies on placing donor atoms at precise distances (20-30 nm) from each other and from control gates.
  • Cavity Coupling: Achieving the necessary weak dispersive interaction requires very high-Q cavities and precise tuning of the coherent light pulses.
  • Integration: Realizing the experiment requires integrating complex cryogenics, high-speed electronics, and long-distance fiber optics into a single cohesive system.

Summary updated

Abstract

So far, all experimental tests of Bell inequalities which must be satisfied by all local realistic hidden-variable theories and are violated by quantum mechanical predictions have left at least one loophole open. We propose a feasible setup allowing for a loophole-free test of the Bell inequalities. Two electron spin qubits of donors31P in a nanoscale silicon host in different cavities 300 m apart are entangled through a bright coherent light and postselections using homodyne measurements. The electron spins are then read out randomly and independently by Alice and Bob, respectively, with unity efficiency in less than 0.7 µs by using optically induced spin to charge transduction detected by radio-frequency single electron transistor. A violation of Bell inequality larger than 37% and 18% is achievable provided that the detection accuracy is 0.99 and 0.95, respectively.
BibTeX
@article{hong2011proposal,
  author = {Fang-Yu Hong and Shi‐Jie Xiong and Yang Xiang and Weihua Tang},
  title = {{Proposal for a Loophole-Free Bell Test with Electron Spins of Donors in Silicon}},
  journal = {Journal of Modern Physics},
  year = {2011},
  doi = {10.4236/jmp.2011.29128},
}
Figure