Stern-Gerlach interferometry in three dimensions: the role of transverse fields

This paper demonstrates that transverse fields, which inevitably accompany static field gradients, cause significant variations in the sensitivity and fringe visibility of three-dimensional Stern-Gerlach interferometers, implying that only specific implementation sequences can achieve high-precision measurements with ultracold Rydberg atoms.

The Gist

Imagine you have a very delicate, magical egg (let's call it "Humpty-Dumpty"). This egg isn't just a normal egg; it's a quantum atom that can exist in two different "moods" or states at the same time.

In a Stern-Gerlach Interferometer (SGI), scientists try to do a magic trick:

1. Split the egg: They use a force to push the "happy" mood of the egg one way and the "sad" mood the other way.
2. The Journey: The two halves travel along different paths.
3. Reunite the egg: They try to bring the two halves back together perfectly so they merge back into one single, coherent egg.

If they do this perfectly, the egg remembers its journey and creates a beautiful interference pattern (like ripples in a pond meeting). This proves the egg was in two places at once. But if the egg gets even slightly "scrambled" during the trip, the magic fails, and the pattern disappears.

The Problem: The "Wobbly Table"

The paper by Meng, Chan, and Martin points out a huge problem with how people have been trying to do this trick.

Imagine you are trying to roll a ball down a long, straight ramp to split it. You want the ball to go straight down the middle. But, because of the laws of physics (specifically how electric or magnetic fields work), you can't just have a straight ramp. The ramp is slightly curved on the sides.

• The Main Force: This is the force pushing the egg forward (the "longitudinal" direction).
• The Side Forces: These are the "transverse" forces. Because the field that pushes the egg forward also pushes it slightly sideways, the egg gets nudged off the center line.

Think of it like trying to walk a tightrope while someone is blowing wind at you from the side. Even if you walk perfectly straight forward, the wind pushes you off balance. In the past, scientists thought, "If we just make the wind strong enough to push the egg forward, the side-wind doesn't matter."

The authors say: "Actually, the side-wind is the reason the trick fails."

The Three Ways to Try the Trick

The authors tested three different "choreographies" (sequences of moves) to see which one is best at keeping the egg from falling apart due to that side-wind. They named them after the shapes the paths make:

1. The "Bell" Sequence (The Clumsy Approach)

• What happens: You push the egg forward, let it drift, and then push it back.
• The Result: It's like trying to balance a broom on your hand while someone is shaking the table. The egg gets pushed far off the center line and its momentum gets messed up.
• The Verdict: Terrible. The "egg" falls apart almost immediately. You need an impossibly tiny, perfect egg (a cloud of atoms smaller than a human hair) to make this work.

2. The "Diamond" Sequence (The Better Approach)

• What happens: You push the egg, flip its mood (swap its state), let it drift, and then push it back.
• The Result: This cancels out some of the momentum errors. It's like walking a tightrope while holding a balancing pole. You are still wobbly, but you don't fall off as fast.
• The Verdict: Okay, but not great. You can use a slightly bigger egg cloud, but it's still very sensitive to the side-wind.

3. The "Bow" Sequence (The Masterpiece)

• What happens: This is the most complex dance. You push, flip, wait, flip again, and push back.
• The Result: This sequence is like a gymnast doing a perfect routine where every wobble is immediately corrected by the next move. The side-wind pushes the egg one way, but the next move pushes it back exactly where it needs to be.
• The Verdict: Amazing. This sequence is so good at correcting the side-wind that you can use a huge cloud of atoms (a whole millimeter wide!) and still get a perfect result.

Why Does This Matter?

Imagine you want to build a super-precise sensor to measure gravity or dark matter.

• If you use the Bell method, you can only use a tiny, weak signal (a few atoms). It's like trying to hear a whisper in a noisy room with a tiny microphone.
• If you use the Bow method, you can use a massive crowd of atoms. It's like using a giant concert hall microphone. You get a much louder, clearer signal, making your measurements incredibly precise.

The Big Takeaway

The paper teaches us that how you arrange the steps matters more than you think.

Two experiments might look identical on paper (splitting an atom and bringing it back), but if you choose the wrong "dance steps" (the sequence of field changes), the invisible side-forces will ruin your experiment. By choosing the right sequence (the "Bow"), you can ignore those annoying side-forces and build much better quantum sensors.

In short: Don't just push the egg forward; choreograph the whole dance so the egg never loses its balance, even when the wind blows.

Technical Summary

1. Problem Statement

The paper addresses a critical limitation in the realization of Stern-Gerlach Interferometers (SGIs). While SGIs are theoretically capable of achieving high sensitivity to acceleration (e.g., gravity $g$) due to large interferometric areas, experimental implementations have struggled with low fringe visibility.

• The Core Issue: Any static magnetic or electric field gradient used to accelerate atoms inevitably generates transverse field components (off-axis fields) due to Maxwell's equations (specifically $\nabla \cdot \vec{E} = 0$ or $\nabla \cdot \vec{B} = 0$).
• The Consequence: For atoms in high-field seeking states (common in Rydberg atom experiments), these transverse fields create an inverted harmonic oscillator potential. This potential pushes atoms away from the symmetry axis, causing the spatial wavepackets of the two interferometer arms to separate in position and momentum.
• The Gap: Previous analyses often focused on longitudinal motion or assumed idealized 1D scenarios. This paper argues that the choice of the experimental sequence (the timing of field gradients and internal state swaps) dramatically influences how these unavoidable transverse fields degrade the interferometric visibility.

2. Methodology

The authors employ a fully quantum mechanical approach to model the SGI, avoiding semi-classical approximations.

• System Model: They analyze a proposed SGI using ultracold Rubidium (Rb) Rydberg atoms accelerated by spatially varying electric fields.
  • States: A superposition of a ground state $|g\rangle$ (negligible dipole moment) and a Rydberg state $|r\rangle$ (large electric dipole moment, $\mu_r \approx 3900 q a_0$).
  • Fields: A cylindrically symmetric electric field configuration where the longitudinal gradient $\partial_z E_z$ drives acceleration, and transverse gradients arise naturally.
• Density Matrix Formalism:
  • The initial state is described by a density operator incorporating both spatial distribution (width $\sigma$) and thermal temperature (de Broglie wavelength $\lambda$).
  • The evolution is calculated using unitary time-evolution operators for three potential types: free particle, linear potential (longitudinal), and inverted harmonic oscillator (transverse).
• Visibility Calculation:
  • The fringe visibility $V$ is derived as the magnitude of the "complex visibility" $V = V_\perp V_\parallel$, where $V_\parallel$ is the longitudinal component and $V_\perp$ is the transverse component.
  • The authors derive closed-form analytical expressions for $V$ by tracing the density matrix through specific pulse sequences.
• Sequences Analyzed: Three distinct SGI sequences are compared:
  1. Bell: Two field gradient reversals, no internal state swaps ($\pi$-pulses).
  2. Diamond: Three field gradient reversals, one internal state swap.
  3. Bow: Two internal state swaps, no field gradient reversals.

3. Key Contributions

• Generalized Visibility Expressions: The authors derive general closed-form expressions for fringe visibilities that account for 3D motion, specifically isolating the impact of transverse fields.
• Sequence-Dependent Sensitivity: They demonstrate that while all sequences may be "closed" in the longitudinal direction (recombining position and momentum), they differ drastically in the transverse direction.
• Scaling Laws: The paper establishes that the decay of transverse visibility $V_\perp$ with time $\tau$ scales differently depending on the sequence:
  • Bell: Scales as $\tau^2$ (Open in both position and momentum).
  • Diamond: Scales as $\tau^4$ (Open in position, closed in momentum).
  • Bow: Scales as $\tau^6$ (Approximately closed in both).
• Optimization Criteria: They identify that the "Bow" sequence is superior for minimizing transverse decoherence, allowing for much larger cloud sizes and longer interaction times compared to other sequences.

4. Results

Using numerical parameters for a proposed Rb Rydberg SGI ($T=100\,\mu\text{K}$, $\tau=1\,\mu\text{s}$, $\partial_z E_z \approx 300 \times 10^3\,\text{V/m}^2$):

• Transverse Visibility ($V_\perp$):
  • For the Bell sequence, $V_\perp$ drops to 0.5 when the cloud radius $\sigma_\perp \approx 4\,\mu\text{m}$.
  • For the Diamond sequence, $V_\perp$ drops to 0.5 at $\sigma_\perp \approx 80\,\mu\text{m}$.
  • For the Bow sequence, $V_\perp$ drops to 0.5 at $\sigma_\perp \approx 1\,\text{mm}$.
• Implication: The Bow sequence allows for a cloud size 1000 times larger than the Bell sequence for the same visibility. This is crucial because larger clouds contain more atoms, reducing quantum projection noise and improving signal-to-noise ratios.
• Longitudinal vs. Transverse: While longitudinal visibility remains perfect ($V_\parallel = 1$) for closed sequences, the transverse visibility is the limiting factor. The "Bow" sequence effectively cancels the leading-order transverse dephasing terms.
• Blue vs. Red States: The analysis shows that while red (high-field seeking) states experience an inverted harmonic potential, blue (low-field seeking) states experience a normal harmonic potential. However, for the specific parameters of the proposed experiment, the visibility differences between red and blue states are negligible, though the scaling laws for the sequences remain the dominant factor.

5. Significance

• Experimental Feasibility: The paper resolves a major bottleneck in SGI implementation. It proves that "putting Humpty-Dumpty back together" (recombining the split wavepackets) is feasible not just by increasing precision, but by choosing the correct pulse sequence.
• Guidance for Future Experiments: The results provide a clear roadmap for designing SGIs for:
  • Precision measurements of gravity ($g$) and field gradients.
  • Tests of the quantum nature of gravity (using nanodiamonds with NV centers).
  • Dark matter detection.
• Universality: Although the example uses electric fields and Rydberg atoms, the derived scaling laws and conclusions apply equally to magnetic field SGIs (using spin states) and other physical systems, provided the field geometry and energy shifts follow similar symmetries.
• Theoretical Insight: The work highlights that in 3D interferometry, "closing" the loop in position and momentum is insufficient if the transverse potential is not properly managed. The "Bow" sequence is identified as the optimal configuration for maximizing interferometric area while maintaining high visibility in the presence of unavoidable transverse fields.