Aspect Experiments

The Aspect experiments were a sequence of optical Bell tests performed in the early 1980s by Alain Aspect, Philippe Grangier, Gerard Roger, Jean Dalibard, and collaborators. They used polarization-correlated photon pairs from atomic cascades to test Bell inequalities more directly than earlier experiments and culminated in a landmark test with time-varying analyzer settings.

Their importance is historical and methodological. They did not close every loophole in the modern sense, but they made the Bell-test program concrete enough that later loophole-free experiments could be understood as refinements of a clear experimental template.

For the general experimental survey, see Bell Inequality Experiments. For the theorem and inequality, see Bell Theorem and CHSH Inequality.

Bell Tests With Photon Pairs

Bell’s theorem had already shown that local hidden-variable models satisfying Bell assumptions obey inequalities violated by quantum mechanics. The early experimental problem was whether real optical correlations could be measured cleanly enough to test those inequalities.

Freedman and Clauser’s 1972 experiment gave an important early violation. Aspect’s program aimed to improve the experimental design in several ways:

• use bright, controlled photon-pair sources;
• measure polarization correlations at several analyzer settings;
• detect both output channels where possible;
• test Bell inequalities with better statistical precision;
• move toward changing analyzer settings while photons were in flight.

The last point was especially important. Bell and others emphasized that static analyzer settings leave a possible locality worry: the fixed settings could have influenced the source or the distant apparatus through an ordinary subluminal signal before a trial. A common cause of the two particles alone is already allowed in a Bell-local model; dependence on the settings is the concern here. Fast changes of settings during flight were designed to make that coordination harder within a local mechanism.

Experimental Arrangement

Aspect’s experiments used photon pairs emitted in a calcium atomic cascade. The cascade produced pairs with correlated polarizations, which were sent in opposite directions toward polarization analyzers. Each run contributed to coincidence counts for chosen analyzer orientations.

The idealized correlation object is the same one used in the CHSH discussion:

$$E(a,b) = \langle A_aB_b\rangle,$$

where $A_a$ and $B_b$ are encoded as $\pm1$ outcomes for analyzer settings $a$ and $b$. In a polarization experiment, the settings are analyzer angles rather than spin directions. Quantum mechanics predicts sinusoidal correlations as the relative angle changes.

The experimental challenge is that photons are not always detected. If a detector registers only a subset of emitted pairs, the observed coincidences may require a fair-sampling assumption. Aspect’s work improved the experimental situation, but it did not remove the detection loophole in the modern high-efficiency sense.

Two-Channel Analyzer Tests

One of the key improvements in the Aspect program was the use of two-channel polarizers. A one-channel polarizer effectively records one output class and loses or ignores the other. A two-channel analyzer sends the two orthogonal polarization components to different detectors, so both binary outcomes can be counted.

The July 1982 two-channel experiment recorded the two orthogonal polarization outputs. Its binary correlations could therefore be estimated from four joint detection classes.

The improvement did not by itself solve all detection problems. A photon might still fail to be detected, and detector efficiencies were far below the levels later needed for loophole-free photonic tests. But two-channel detection reduced the mismatch between the theoretical Bell scenario and the laboratory coincidence-counting procedure.

Time-Varying Analyzer Settings

The December 1982 time-varying experiment was a separate arrangement. Each acousto-optical switch sent light toward one of two linear polarizers. The switches operated at different, incommensurate frequencies near 50 MHz, changing the effective setting during flight. This was rapid switching between optical paths, not a polarizer mechanically rotating at that frequency.

The logic was spacetime-based. If the setting on one side is changed late enough, then information about that setting cannot reach the other side by any signal traveling at or below the speed of light before the distant outcome is registered. In an ideal Bell test, the relevant setting choices and measurement events are spacelike separated.

Aspect’s switching was a major step toward that ideal, but it was not identical to modern random, event-by-event setting choice. The settings were varied by a periodic switching mechanism, and later experiments used faster random choices and more refined timing analysis. The historical significance remains substantial: the experiment directly addressed a locality concern that earlier static tests left more open.

Results

The experiments found agreement with the quantum correlations and violations of their respective Bell inequalities under the stated sampling assumptions. The December 1982 time-varying paper reported a violation by five standard deviations. Different papers used differently defined Bell parameters: its quoted result should not simply be relabeled as a measurement of the CHSH quantity $S$ defined on this site.

The important result was not merely that the photons were correlated. Classical correlated sources can produce correlations. The Bell-test claim is that the pattern of correlations across several analyzer settings violates a bound satisfied by local hidden-variable models.

The experiments therefore strengthened the empirical case against local hidden-variable explanations of EPR-Bohm correlations. They also made clear that experimental foundations required precision engineering, not only conceptual argument.

Remaining Loopholes

The Aspect experiments were not loophole-free by current standards. The main remaining issues were:

• Detection efficiency was too low to avoid fair-sampling assumptions.
• The time-varying settings were not generated by modern fast random choices.
• The locality arrangement improved on static tests but did not close every relevant spacetime condition in the strongest later sense.
• Coincidence identification and analysis assumptions still mattered.

The later Loophole-Free Bell Tests combined efficient detection and spacelike separated random settings in the same experiment. Comparing these requirements explains both Aspect’s advance and the work still needed afterward.

Exercises

1. Why did two-channel analyzers make the Aspect experiments closer to the ideal Bell-test scenario?

Solution

A Bell-test correlation is built from two outcome classes on each side. A two-channel analyzer records both orthogonal polarization outputs rather than only one selected channel. This makes the experiment closer to assigning a binary outcome to each detected photon, though it does not by itself remove losses from imperfect detection.

2. Explain why changing analyzer settings while photons are in flight addresses a locality concern.

Solution

If a setting is chosen or changed late enough, then information about that setting cannot reach the distant measurement station by any signal traveling at or below the speed of light before the distant outcome is recorded. This reduces the possibility that ordinary subluminal communication coordinates the two outcomes.

3. Why were the Aspect experiments not loophole-free in the modern sense?

Solution

The main limitations were detection efficiency and setting choice. The experiments still relied on fair-sampling assumptions because not every emitted pair was detected, and the time-varying settings were generated by a periodic switching mechanism rather than modern fast random event-by-event choices. Later experiments addressed these issues more directly.

4. Why does a Bell violation require several analyzer settings rather than only observing strong same-angle correlations?

Solution

Strong same-angle correlations can be produced by ordinary correlated sources. A Bell violation concerns the pattern of correlations across incompatible setting pairs. The CHSH combination uses four correlations with a fixed sign pattern, and it is this multi-setting pattern that can exceed the local hidden-variable bound.

References

• A. Aspect, “Proposed experiment to test the nonseparability of quantum mechanics,” Physical Review D 14, 1944-1951, 1976, DOI: 10.1103/PhysRevD.14.1944.
• A. Aspect, P. Grangier, and G. Roger, “Experimental Tests of Realistic Local Theories via Bell’s Theorem,” Physical Review Letters 47, 460-463, 1981, DOI: 10.1103/PhysRevLett.47.460.
• A. Aspect, P. Grangier, and G. Roger, “Experimental Realization of Einstein–Podolsky–Rosen–Bohm Gedankenexperiment: A New Violation of Bell’s Inequalities,” Physical Review Letters 49, 91-94, 1982, DOI: 10.1103/PhysRevLett.49.91.
• A. Aspect, J. Dalibard, and G. Roger, “Experimental Test of Bell’s Inequalities Using Time-Varying Analyzers,” Physical Review Letters 49, 1804-1807, 1982, DOI: 10.1103/PhysRevLett.49.1804.
• J. S. Bell, Speakable and Unspeakable in Quantum Mechanics, 2nd ed., Cambridge University Press, 2004.
• N. Brunner, D. Cavalcanti, S. Pironio, V. Scarani, and S. Wehner, “Bell nonlocality,” Reviews of Modern Physics 86, 419-478, 2014, DOI: 10.1103/RevModPhys.86.419.
• The Nobel Prize in Physics 2022, NobelPrize.org, award summary.