Entangling ions with engineered light gradients

Introduction to the show: ident: Quantum Radio. Generated commentary on the latest quantum physics and condensed matter papers.

Kai: We're looking at the paper: "Entangling ions with engineered light gradients".

Mira: Here's the gist of the paper. A geometric-phase entangling interaction driven by a transverse, time-dependent structured-light force reduces spectral crowding in trapped-ion quantum processors while preserving single-ion addressing, enabling high-fidelity two-qubit gates.

Kai: Spectral Crowding and the Problem Spectral crowding of collective motional modes limits the fidelity of entangling interactions in trapped-ion quantum processors by inducing off-resonant coupling to spectator modes This issue arises across multiple platforms, including silicon quantum dots

1: , nitrogen-vacancy centres in diamond

2: and superconducting transmon processors

3, 4: , where off-resonant excitation of unwanted transitions can degrade performance In trapped-ion systems, ions are confined by static and radio-frequency electric fields, forming linear crystals whose collective motion can be described in terms of quantised normal modes However, as the number of ions increases, the frequency separation between adjacent motional modes decreases Entangling gates that rely on spectrally addressing one or a small subset of these modes therefore suffer from parasitic coupling to nearby spectator modes, limiting coherent control and scalability The Proposed Geometric Phase Interaction The paper introduces an alternative approach to suppressing off-resonant motional couplings by engineering the spatial structure of the driving field rather than relying on dynamical decoupling They implement a light-shift (LS) entangling gate based on electric field gradients, generated by a tightly focused laser beam in a TEM10-like transverse mode By interfering optically structured beams, they create a state-dependent optical dipole force aligned with the axis of a linear ion crystal, while the beam propagates transversely to it This configuration enables targeted addressing of selected ion pairs with minimal crosstalk and mediates a conditional geometric phase interaction through the axial motional spectrum, operating in a spectral region that is less susceptible to coupling to non-targeted modes Mechanism of Entanglement Generation The entangling interaction leverages the properties of axial motional modes which are more widely spaced than radial modes in one-dimensional harmonically-confined ion chains The two beams generate an oscillating transverse AC Stark shift gradient at the ions’ position (Fig.

Mira: 1-b) If the beams exhibit lin∥ lin polarisation, the AC Stark shift gradient is induced by an intensity gradient and is differential for states separated by an optical transition Conversely, if the beams exhibit lin⊥ lin polarisation, the AC Stark shift gradient is induced by a polarisation gradient and is differential for states with different orbital angular momentum This gradient induces an optical dipole force (ODF), which can be used to generate entanglement via the accumulation of statedependent geometric phases The two-qubit gate implements a σz ⊗ σz interaction, which can be generalised to an arbitrary σϕ = σx cos ϕ + σy sin ϕ interaction with local single-qubit rotations Experimental Realization and Performance The experimental setup involves a chain of 138Ba+ ions loaded into a linear Paul trap with secular trap frequencies (ω(x), ω(y), ω(z)) = 2π × (1.304, 1.344, 0.3422) MHz Qubits are encoded in the 0⟩ = mj = −1/2⟩ and 1⟩ = mj = +1/2⟩ Zeeman sublevels of the GS The ions are first cooled to the Doppler limit (n¯(ax)COM ≈ 22) via the 6S1/2 → 6P1/2 transition The two-qubit gate consists of four phase-space loops at δ = 2π × 20 kHz, with a tgate = 4 · (2π/δ) ≈ 220 µs The observed infidelities lie below the fault-tolerance threshold, enabling the encoding of logical qubits and the implementation of quantum error correction (QEC) protocols Error Analysis and Conclusion The dominant error source limiting fidelity is parasitic coupling to axial spectator modes, with an infidelity in the order of 10−4 In contrast, performing the same gate on radial modes would yield an infidelity of 7.

Title and authors: Kai: So we're looking at this paper, "Entangling ions with engineered light gradients," which deals with how to do two-qubit gates in trapped ions without messing up the other modes. It seems like they’re trying to fix a problem where too many motional modes get excited when you try to entangle just one pair.

Mira: They are tackling that spectral crowding issue head-on, and they suggest using a transverse structured-light force instead of relying on dynamical decoupling sequences to manage those off-resonant interactions.

The paper's summary: Kai: So, we’re looking at how they took that idea about structured light and turned it into something actually working for two-qubit gates in trapped ions, page eight, they talk about the core idea.

Mira: Basically, they showed you can use this engineered force—this transverse structured-light force—to create a geometric phase interaction.

Kai: Right, so instead of just flipping a switch to change the entanglement dynamics dynamically, they’re building this state-dependent optical dipole force using electric field gradients from a tightly focused laser beam.

Mira: That gradient is what makes the interaction conditional; it means the force you get depends on which qubit states you are in.

Kai: And because they set up this light propagation sideways to the ion chain, they manage to target specific ion pairs with very little crosstalk, which is a big deal for scaling up.

Mira: They focus this entire process onto the axial motional spectrum, which is where their modes are more spread out than the radial ones, and that’s why it works better.

Kai: The result they report is that they achieved an infidelity in the order of one hundred thousand, ten-four when coupling to those axial spectator modes.

Mira: That number really puts it in perspective because if they tried to do the same thing on radial modes, the infidelity jumped up to seven, which shows how much cleaner this axial approach is.

Kai: Lev needs to look at that number because for a real system with twelve ions, you need those error rates down below five times ten negative three just to start thinking about fault tolerance.

Mira: So what this means for us is they’ve found a way to suppress the dominant error source in these systems—the coupling to unwanted modes—by engineering the light field itself instead of relying on post-processing or dynamical decoupling sequences.

Kai: That’s the core implication, moving away from just trying to fix errors after they happen, and instead building them into the interaction itself.

Mira: It suggests that spatial light modulators aren't just for fancy imaging; they can be a fundamental tool for designing high-fidelity quantum gates in these hardware platforms.

The paper's improvements: Kai: They specifically look at scaling high-fidelity entanglement to bigger ion chains using these gradient-field light-shift gates.

Mira: That’s the point; if you use this method, you can maintain error rates below five times ten negative three even when you push it out to twelve ions, which is a much more realistic number for building a scalable processor.

Kai: So that means they’ve found a way to keep the fidelity high enough for logical qubits and even start thinking about quantum error correction protocols on larger hardware.

Mira: And they also talk about making the control more flexible; they suggest using a spatial light modulator instead of just a static phase plate for controlling those force vectors.

Kai: That means you can program different gradient orientations, which could allow you to do single- or multi-ion entangling interactions even when you have ions arranged in two dimensional geometries.

Mira: It’s about making the hardware platform itself more versatile so it isn't locked into one fixed addressing configuration for every possible gate.

Kai: Lev would be interested in that flexibility because programmable control could reduce the complexity of the classical control electronics needed to drive all those different light patterns.

Mira: And if they can program those vectors, it opens up new ways to design gates that might work better for different ion arrangements or connectivity requirements.

Kai: So, they’re not just showing a proof of concept for a small chain; they're laying out a path to build a more robust system where you can scale both the number of ions and the complexity of the gates.

Conclusion: Kai: So to wrap things up, we’re talking about "Entangling ions with engineered light gradients," and the main thing is they successfully used structured light to create a geometric phase interaction that specifically targets axial motional modes for two-qubit gates.

Mira: That’s right, it means they managed to suppress the parasitic coupling to spectator modes by engineering the spatial structure of the driving field instead of relying on dynamical decoupling sequences.

Kai: They showed this works experimentally with one hundred thousand infidelity when they were operating in that axial spectral region, which is a big step because those modes are more widely spaced.

Mira: The implication for us is that we can design interactions based on motional mode spacing instead of just brute-forcing dynamical decoupling sequences to manage noise.

Lev: For real hardware, this means the overhead for running these gates should land much lower than if we had to fight those off-resonant couplings dynamically across the entire spectrum.

Kai: Exactly, it's about making the interaction itself more robust by using geometry rather than just timing pulses to fix imperfections.

Mira: It also opens up avenues for building more complex architectures where you might need to address different pairs in different ways because of that programmable force vector they discussed.

Lev: If we can program those vectors, it cuts down on the classical control complexity because you’re not just running a fixed sequence anymore.