Measuring the spin of a spin-1/2 and getting a result of 7

AI summary

Overview

Research area: Quantum measurement, specifically weak values and weak value amplification (WVA) in atomic systems; AMO physics with Bose-Einstein condensates.

Technical level: Advanced. The paper assumes familiarity with weak measurement formalism, von Neumann measurement models, and atomic hyperfine structure (Breit-Rabi formula, rf/microwave state manipulation).

Scope (one sentence): The paper reports the first atomic realization — and the closest realization to date of the original 1988 Aharonov-Albert-Vaidman proposal — of a weak Stern-Gerlach experiment, using a ^87^Rb Bose-Einstein condensate to amplify a small magnetic momentum kick by a factor of 14 through unlikely postselection, yielding an anomalous weak value of 7(3) for an effective spin-1/2.

What This Paper Is About

When a quantum system is measured weakly and then postselected on a final state that nearly cancels the initial state, the average pointer reading can be amplified far beyond the range allowed by the operator's eigenvalues — an "anomalous weak value." This effect has been seen in light, superconducting qubits, and neutrons, but the specific example used to introduce the concept in 1988 — weakly measuring a component of a particle's spin in a Stern-Gerlach geometry — had never been performed as proposed, largely because matter waves have very short coherence lengths. The authors use a BEC of ^87^Rb, whose coherence length is far larger than that of a thermal atomic beam, to perform this weak Stern-Gerlach measurement and observe the resulting amplification.

Key Contributions

1. First atomic weak value amplification. The authors demonstrate WVA using atoms, which they describe as the first anomalous weak value observed using atoms and the first anomalous measurement of a particle's spin with a true continuous probe.
2. Closest realization of the original 1988 proposal. The experiment implements a weak Stern-Gerlach measurement of a spin component with a continuous pointer (momentum), the scenario Albert, Aharonov, and Vaidman used to introduce anomalous weak values. The paper notes this class of experiment had previously been called "infeasible" in the literature.
3. A novel weak-coupling mechanism. Because both states used (|2,0⟩ and |1,0⟩) are m = 0 states, a magnetic gradient produces no first-order force; instead the authors exploit the second-order Zeeman shift (Breit-Rabi) to generate equal and opposite momentum kicks, functioning identically to a spin-1/2 in a standard Stern-Gerlach experiment.
4. Demonstration that increasing the kick weakens the amplification. The paper shows experimentally, and supports with an analytic Gaussian model, that a larger momentum kick makes the measurement "less weak," causing earlier deviation from the weak value prediction and a smaller maximum amplification factor.

Main Findings

• Measured amplification factor of 14. Postselecting on a nearly orthogonal state produced a momentum shift 14 times larger than the shift actually applied by the gradient, for a kick of δp = 4.4(5) µm/s (the abstract characterizes the kick as ∼8 µm/s).
• Anomalous weak value of 7(3). The largest pointer shift, at δ = 0.38, was 14(6) δp from zero — 7(3) times larger than the shift observed at δ = ±π/2 — corresponding to a weak value of 7(3) for σ_z, as though a measurement of the z component of the effective spin-1/2 gave m_z = 7, far outside the ±1/2 eigenvalue range.
• Theory matches the shape of the data. The measured shifts exceed 2δp in magnitude on either side of δ = 0 and pass through zero at δ = 0, consistent with the weak value prediction. The width of the region where the weak approximation breaks down agrees well with the analytic prediction; the observed scale of amplification appears to slightly exceed theory.
• Larger kick, weaker amplification. With a roughly 2× larger kick (δp = 9.2 ± 0.8 µm s⁻¹), the data deviate from the weak value prediction sooner and over a larger range of δ, and the maximum amplification factor is smaller.
• Amplification factor passes through zero, not infinity. Although the weak value diverges at δ = 0, the amplification factor π/δp goes through zero there, a breakdown of the weak approximation discussed by Duck and Sudarshan; the region of breakdown can be made arbitrarily narrow as δp → 0 but never vanishes.
• Limits on the maximum shift. For large δp, the pointer shift is bounded by ±δp (no amplification); for small δp the maximum shift becomes independent of δp and equals 1/(σ√2), of order 1/σ.
• A coherence-momentum argument. The relevant comparison scale for "large" or "small" δp is not the momentum width Δp but ħ/Δx, a "coherence momentum." Only in a Heisenberg-limited state do these coincide; a BEC makes the coherence momentum large enough to observe interference.
• Attempted smaller-cloud test did not succeed. The authors expected a smaller cloud to yield higher, narrower amplification peaks but could not observe this, because the only available method to shrink the cloud also reduced atom number and lost signal.

Methodology in Plain English

The team starts with a BEC of about 5-10 × 10³ ^87^Rb atoms in a crossed optical dipole trap: one "waveguide" beam giving weak longitudinal confinement (ω = 3.7 Hz) and tight radial confinement (ω ≈ 300 Hz), plus a second beam for longitudinal confinement. The cloud's longitudinal size ranges from 18 to 33 µm depending on atom number.

They prepare the atoms in |F=2, m_F=2⟩, transfer them to |2,0⟩ with two-photon RF adiabatic rapid passage, and then "shelve" away any stray population in other m levels using a microwave π pulse to |1,0⟩ followed by 0.5 ms of light resonant on the F=2 → F′=3 D2 transition, which ejects any remaining F=2 atoms from the trap.

A π/2 pulse then creates an equal superposition of |2,0⟩ and |1,0⟩ — the initial state, analogous to |+x⟩ for a spin-1/2. A magnetic gradient pulse provides the weak coupling. Since both states have m = 0, the coupling comes from the second-order Zeeman effect described by the Breit-Rabi formula (with ΔE₀ = 6.835 GHz and nuclear spin I = 3/2); the resulting force depends on both the gradient and the local field magnitude, so a transverse field component matters too. Field strengths are calibrated by Ramsey spectroscopy to about 10 Hz precision, corresponding to mG-level field uncertainty.

Rather than let the cloud expand freely (which would cost too much density and thus imaging precision), they release the atoms into the waveguide. A harmonic oscillator rotates states in phase space, so after one quarter of a period the momentum kick becomes a position shift.

Postselection is done with a variable-duration microwave pulse on the m = 0 transition that maps a chosen final state onto the imaged state |2,0⟩. They image both the postselected state |f(δ)⟩ and the orthogonal state |n(δ)⟩, fitting each absorption image with a 2D Thomas-Fermi profile and subtracting the two centers. This common-mode subtraction gives error bars 5-10 times smaller than subtracting averages of first and second images separately, and also yields the postselection probability and hence the overlap ⟨f|i⟩ shot by shot. The ARP transfer adds 5.6 ms of expansion time between the two images, which is accounted for. Each data point averages 60 image pairs, with error bars giving the standard error on the mean.

Images whose fitted centers fell outside the extreme edges of high-density clouds at δ = ±π/2 were discarded, justified because the postselected cloud density is everywhere lower than that of the interfering clouds and must vanish outside that range. Such spurious points arise from camera dark-count shot noise when atom number fluctuations drop the density too low, and they broaden the distribution and inflate the error in the mean.

Why This Matters

Impact on research: The experiment closes a long-standing gap between the original pedagogical proposal for weak values and its physical realization, showing that anomalous weak values are accessible in atomic systems despite their small coherence volumes. It also demonstrates that second-order Zeeman coupling between two m = 0 hyperfine states can act as a clean, tunable spin-momentum coupling — functionally identical to a spin-1/2 Stern-Gerlach interaction — which is a technique others could borrow.

Real-world applications (as suggested by the authors):

• Magnetic gradiometry, where amplification could make small field differences more visible.
• Searches for small magnetic moments, where weak value amplification could enhance a tiny signal.
• The paper sits in a broader context of proposals it cites for WVA in neutron electric dipole moment measurements and for an enhanced version of LIGO.
• Sensing at very low velocities: the technique "measured" a momentum shift on the µm/s scale, a rarely accessed velocity regime.

Industry relevance: Quantum sensing and precision metrology, particularly magnetometry and interferometry, are the most direct beneficiaries. The authors caution that any practical metrology application would likely require improving signal-to-noise and refining the technique, since the amplified points with unlikely postselection carry large errors.

Future Directions

• Improve signal-to-noise. The errors on the highly amplified, low-postselection-probability points are currently too large to be useful for metrology; reducing them is a prerequisite for any sensing application.
• Use a Feshbach resonance. The authors propose using an atomic species with a convenient Feshbach resonance to reduce interaction energy before the momentum kick, producing a smaller cloud after expansion and higher density for a given atom number — which would also allow the untested prediction of higher, narrower amplification peaks from a smaller cloud to be checked.
• Push toward smaller momentum kicks and stronger amplification. The theory predicts the breakdown region narrows and the amplification factor approaches the weak value as δp → 0; reaching that regime experimentally remains open.
• Apply the technique to small magnetic moments and gradiometry, the authors' stated long-term motivation, contingent on the sensitivity improvements above.

Target Audience

Quantum measurement theorists and experimentalists working on weak values, foundations of quantum mechanics, and precision metrology; AMO physicists interested in BEC manipulation, hyperfine state control, and spin-momentum coupling via the second-order Zeeman effect; and graduate students or advanced undergraduates looking for a concrete, experimentally realized example of an anomalous weak value with matter waves. Readers without background in weak measurement formalism or atomic hyperfine structure will find the theory and methods sections demanding.

Authors’ abstract

The weak value of an operator is the average result of measuring that operator on a quantum system given specified initial and final states. The overlap between these states appears in the denominator of the weak value, and thus if that overlap is small, the weak value can be arbitrarily large. In 1988, Albert, Aharonov, and Vaidman proposed a weak Stern-Gerlach experiment which would be able to measure a so-called anomalous weak value, a measurement result which lies out- side of the eigenvalue spectrum of the operator being measured. Though anomalous weak values have been measured in numerous optical systems, here we present the first such realization of this experiment as originally proposed. By placing a Bose-Einstein condensate of 87Rb in a magnetic gradient and performing an unlikely postselection, we measure an effective magnification of a weak magnetic momentum kick (~ 8 $μ$m/s) by a factor of 14. We also demonstrate that increasing the momentum kick decreases the "weakness" of the measurement and reduces the amplification.