An efficient method to generate near-ideal hollow beams of different shapes for box potential of quantum gases

This paper presents a hybrid optical scheme combining fixed optics (axicons and prisms) with a digital micromirror device to efficiently generate high-quality, near-ideal hollow beams of various shapes, enabling the creation of highly homogeneous cold atom samples in customizable box potentials for advanced quantum many-body physics research.

The Gist

Imagine you are a chef trying to bake a perfect, flat cake for a very delicate dessert. In the world of quantum physics, scientists are trying to do something similar: they want to trap a cloud of super-cold atoms in a perfectly flat, uniform "box" so they can study how these atoms behave together.

The problem is that the tools they usually use are like baking in a bowl. The bowl (a traditional trap) forces the ingredients (atoms) to pile up in the middle and thin out at the edges. This makes it hard to study them fairly because the "temperature" and "pressure" change depending on where you look.

To fix this, scientists want to build a square or round "box" made of light that has perfectly straight, vertical walls and a completely flat floor. This is called an "optical box trap." But building a light box with perfect walls is incredibly difficult.

The Old Ways: The "Cookie Cutter" vs. The "Sieve"

Previously, scientists tried two main methods to make these light boxes, and both had flaws:

1. The Fixed Cookie Cutter (Fixed Optics): They used special glass lenses (like axicons) to bend a laser beam into a ring or a square.
   • The Flaw: It's like using a cookie cutter that you can't change. If you want a different shape, you need a whole new cutter. Also, these cutters often leave behind "crumbs" (stray light) in the middle of the box, which ruins the flatness of the floor.
2. The Digital Sieve (DMDs): They used a Digital Micromirror Device (DMD), which is like a screen made of millions of tiny mirrors that can flip on and off to block light.
   • The Flaw: To make a hollow box, they had to take a solid beam of light and block out the middle. This is like trying to make a donut by throwing away the center of a bagel. You lose a huge amount of your "dough" (laser power), making the box weak and inefficient.

The New Solution: The "Pre-Shape and Polish" Method

The team in this paper came up with a clever hybrid approach that combines the best of both worlds. Think of it as a two-step cooking process: Pre-shaping and Polishing.

Step 1: The Pre-Shape (The "Mold")

First, they use the "fixed optics" (the special glass lenses or prisms) to take a solid laser beam and bend it into a rough hollow shape (like a ring or a square).

• Analogy: Imagine you have a ball of clay. Instead of trying to carve a hollow bowl out of it with a knife (which wastes clay), you use a special mold to press it into a rough bowl shape immediately. Most of the clay stays in the walls of the bowl; you haven't thrown much away yet.

Step 2: The Polish (The "Digital Sculptor")

Now, the beam is roughly the right shape, but it's messy. There is still some stray light in the very center and fuzzy edges. This is where the DMD (the digital mirror screen) comes in.

• Analogy: Instead of using the DMD to carve the whole shape from scratch (which wastes power), they use it like a digital sculptor or a high-precision eraser. It gently sweeps away the tiny bits of stray light in the center and sharpens the fuzzy edges to make them razor-thin.

Why is this a Big Deal?

The results of this new method are impressive:

1. Super Sharp Walls: The walls of their light box are incredibly steep. If you were to draw a graph of the wall's height, it would look like a cliff rather than a ramp. In math terms, the "steepness" is over 100 times sharper than previous attempts. This means the atoms are trapped in a very strict, uniform box.
2. No Waste: Because they pre-shaped the beam first, they didn't have to throw away most of the laser power. They are about 3 times more efficient than the old methods. This means they can use weaker, cheaper lasers and still get a strong trap.
3. Shape-Shifting: Because the DMD is programmable, they can change the shape of the box on the fly. Want a ring? Done. Want a square? Done. Want a pentagon? Just change the computer code. No need to buy new glass lenses.

The Bottom Line

This paper describes a new "recipe" for trapping cold atoms. By using a mold to get the general shape and a digital eraser to clean up the details, the scientists have built a nearly perfect, flat, and uniform box of light.

This tool allows physicists to study quantum gases (like super-cold atoms) in a way that was previously impossible, potentially leading to breakthroughs in understanding how matter behaves at the smallest scales, which could help build better quantum computers and sensors in the future.

Technical Summary

1. Problem Statement

Ultracold quantum gases are typically prepared in harmonic traps, which suffer from atomic density inhomogeneity. This leads to position-dependent energy, length scales, and temperature, complicating the manipulation and detection of the system. Furthermore, non-local quantities (e.g., correlation functions) are often smeared out in trap-averaged measurements.

While optical box traps (using blue-detuned hollow beams) offer a solution by creating uniform potentials, existing methods have significant drawbacks:

• Fixed Optics (Axicons/Phase Plates): Generate hollow beams but often require non-adjustable masks. They struggle to eliminate central stray light and residual fringes, leading to imperfect box potentials.
• Programmable Spatial Light Modulators (e.g., DMDs): Offer flexibility but typically shape Gaussian beams by directly cutting away the central portion. This results in low optical efficiency (wasting laser power) and limits the achievable trap depth due to laser power constraints and DMD damage thresholds.
• Edge Quality: Previous 3D box traps had low potential steepness (power-law exponents < 17), and 2D ring traps often required expensive, custom microscopic objectives.

2. Methodology

The authors propose a hybrid scheme combining fixed optics for pre-shaping and a Digital Micromirror Device (DMD) for fine-tuning to generate near-ideal hollow beams.

• Stage 1: Pre-shaping with Fixed Optics

  • Ring Beams: A pair of axicons (conical lenses) with identical cone angles are placed tip-to-tip on the optical axis. This reshapes the input Gaussian beam into a Bessel-like near-field beam and a collimated ring-shaped beam in the far field.
  • Square Beams: Two pairs of large-angle multi-faceted prisms are used to split and recombine the beam, creating a pre-shaped square hollow beam.
  • Outcome: These optics redistribute most of the laser power into the hollow region, leaving only residual light in the center and around the edges.

• Stage 2: Refinement with DMD

  • A Digital Micromirror Device (DMD) is placed in the optical path to act as a programmable mask.
  • The DMD is programmed to reflect the "ON" state pixels (forming a solid circle or square) to match the pre-shaped beam, while the "OFF" state pixels deflect the residual central light and inner fringes away from the optical axis.
  • This step eliminates stray light and sharpens the inner boundary without sacrificing the bulk of the laser power.

• Stage 3: Imaging System

  • A high-resolution imaging system using a high Numerical Aperture (NA) microscope objective (Mitutoyo, 50X, NA=0.5) and a plano-convex lens demagnifies the beam by a factor of 62.5.
  • This projects the beam onto the atom position with high spatial resolution (~0.67 µm).

3. Key Contributions

• Hybrid Architecture: Successfully integrates fixed optics (for high-efficiency power redistribution) with programmable DMDs (for high-quality edge definition).
• Shape Versatility: Demonstrates the ability to generate both ring-shaped and square-shaped hollow beams, with the potential for other geometries (pentagon, hexagon) by swapping fixed optical elements.
• Real-time Optimization: The DMD allows for real-time alignment and correction of the beam profile, enabling automated calibration protocols.

4. Key Results

• Potential Steepness (Edge Quality):
  • The intensity profile of the inner beam edge was fitted to a power-law function $I(x) = Ax^\alpha$.
  • Ring Beam: Achieved an exponent of $\alpha = 80 \pm 7$.
  • Square Beam: Achieved an exponent of $\alpha = 104 \pm 9$.
  • Significance: These values represent the steepest potential walls reported to date (previous records were < 17 for 3D and required custom optics for 2D). The inner edge width is measured at 0.93 µm.
• Light Efficiency and Peak Intensity:
  • The authors define a figure of merit as the product of peak intensity ($I_{peak}$) and efficiency ($\eta$).
  • Compared to direct DMD shaping or masking of Gaussian beams, the hybrid method improves the $I_{peak} \cdot \eta$ product by a factor of ~7.3 (at a ring radius ratio of 0.85).
  • The method preserves the peak intensity because the fixed optics redistribute power to the hollow region before the DMD removes only the small residual center.
• Residual Light:
  • After DMD optimization, the average intensity noise from residual light in the center is < 2% of the peak intensity, ensuring high density homogeneity.

5. Significance

• Quantum Simulation Platform: This method provides a versatile platform for creating homogeneous quasi-2D quantum gases with different geometric boundaries.
• Advanced Physics Exploration: The high uniformity and steep walls enable unprecedented studies of:
  • Quantum many-body physics.
  • Dynamical phase transitions.
  • Berezinskii–Kosterlitz–Thouless (BKT) physics.
  • Sound propagation and fermionic pairing in uniform systems.
• Practical Advantages: By significantly improving light efficiency, the scheme reduces the demand on laser power and lowers the risk of damaging sensitive optical components like DMDs, making high-quality box traps more accessible for experimentalists.

In conclusion, this work presents a robust, efficient, and flexible solution for generating optical box potentials, overcoming the trade-offs between beam quality, efficiency, and programmability that have limited previous approaches.