🔭 astrophysics

Asgard/NOTT: L-band nulling interferometry at the VLTI -- III. The mid-infrared integrated optics beam combiner for NOTT

This paper reports the successful manufacturing and laboratory characterization of a mid-infrared, four-telescope integrated optics beam combiner for the NOTT instrument, demonstrating a record-breaking self-calibrated deep null of $1.14 \times 10^{-3}$ in the L' band using a double-Bracewell architecture at room temperature.

Original authors: A. Sanny, L. Labadie, S. Gross, K. Barjot, R. Laugier, G. Garreau, M. -A. Martinod, D. Defrère, M. J. Withford

Published 2026-04-20

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: A. Sanny, L. Labadie, S. Gross, K. Barjot, R. Laugier, G. Garreau, M. -A. Martinod, D. Defrère, M. J. Withford

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to hear a tiny, whispering firefly (a young planet) sitting right next to a blindingly bright stadium floodlight (a star). In the universe, this is the ultimate challenge for astronomers: finding planets that are very close to their stars. The light from the star is so overwhelming that it completely drowns out the faint glow of the planet.

This paper describes a new, high-tech "noise-canceling headphone" for telescopes, designed specifically to solve this problem.

The Problem: The Star is Too Loud

Astronomers want to study planets near the "water snowline"—a specific distance from a star where water turns to ice. This is a sweet spot where giant planets like Jupiter often form. However, looking at these planets is like trying to see a candle flame next to a laser pointer. The star's light is millions of times brighter than the planet's.

The Solution: The "NOTT" Headphones

The team built a device called NOTT (Nulling Observations of dust and planeTs) to attach to the VLTI, a massive telescope array in Chile. Think of NOTT as a sophisticated pair of noise-canceling headphones for light.

Instead of just blocking the star's light (which is hard because the star and planet are so close together), NOTT uses a trick called interferometry. It takes the light coming from four different telescopes and mixes them together in a very specific way.

The Analogy: Imagine four people shouting in a room. If they all shout at the exact same time, the noise is loud. But if you arrange them so that two people shout "Up" while the other two shout "Down" at the exact same moment, the sound waves cancel each other out. The result? Silence.
The Magic: NOTT arranges the light waves from the star so they cancel each other out (creating "silence" or a "null"). However, because the planet is in a slightly different spot, its light doesn't cancel out. It survives the cancellation, allowing the telescope to finally "hear" the planet.

The New Chip: The "Brain" of the Headphones

The core of this paper is about the brain inside these headphones: a tiny glass chip called an Integrated Optics Beam Combiner.

The Material: They made this chip out of a special glass called Gallium Lanthanum Sulfide. Think of this glass as a super-highway for light that works perfectly in the "infrared" part of the spectrum (heat light), which is where young planets glow the brightest.
The Manufacturing: They didn't use a saw or a drill. Instead, they used Ultrafast Laser Inscription. Imagine using a super-fast, microscopic laser pen to "draw" tiny tunnels inside the glass. These tunnels guide the light from the four telescopes to the mixing point.
The Design: The chip uses a clever pattern called the "Double-Bracewell" architecture. It's like a traffic system with four lanes that merge and split in a specific dance.
- Two lanes merge, cancel out the star's light, and then merge again with the other two lanes.
- This creates a "broad null," meaning it cancels out the star's light over a wider area, making it easier to spot the planet even if the telescope isn't perfectly aligned.

The Results: A Successful Test Run

The team built this chip and tested it in a lab at room temperature (not yet in the freezing cold of space or a cryogenic chamber). Here is what they found:

It Works: They successfully created a "null" (silence) where the star's light was reduced by a factor of 1,000 to 10,000. This is a huge success for a first-generation device.
The "Self-Calibration" Trick: The chip is smart. It has a built-in way to check its own work. By comparing two different "silenced" outputs, it can mathematically remove errors caused by the atmosphere or the telescope itself. This is like the headphones automatically adjusting to the shape of your ear to cancel noise better.
Efficiency: About 37% of the light gets through the chip. The authors say that if they add a special anti-reflective coating (like the coating on camera lenses), this could jump to nearly 50%.
Polarization: They checked if the glass twisted the light in weird ways (like a pair of sunglasses that only work at a specific angle). They found the twisting was very small, which is good news.

What's Next?

This paper is the "prototype" phase. The chip works great in the lab, but the next step is to put it in a freezer (cryogenic conditions). Why? Because heat creates "noise" (thermal background) that can hide the faint planet signal. Cooling the chip down will make the "silence" even deeper.

In Summary:
This paper is a report card for a new, tiny glass chip that acts as a super-smart light mixer. It successfully proved that we can use lasers to "draw" tunnels in glass that cancel out the blinding light of stars, giving us a clear view of the hidden, young planets orbiting them. It's a major step toward finding out how planets like our own (or even bigger ones) are born.

1. Problem Statement and Context

The detection and characterization of young, self-luminous giant exoplanets and hot exozodiacal dust near the water-ice line require high-contrast imaging in the mid-infrared (specifically the $L'$ band, 3.5–4.1 $\mu$ m). Current direct detection techniques face challenges due to the extreme contrast between the planet and its host star ( $10^{-3}$ to $10^{-6}$ ) and small angular separations (1–10 mas).

Nulling interferometry is the prime technique to overcome these limits by destructively interfering starlight while transmitting off-axis planetary light. The NOTT (Nulling Observations of dust and planeTs) instrument, a visitor instrument for the VLTI (Very Large Telescope Interferometer), aims to achieve this using a four-telescope (4T) integrated optics (IO) beam combiner.

The specific technical challenge addressed in this paper is the development and characterization of the first mid-infrared, four-telescope IO beam combiner based on the double-Bracewell architecture. This device must operate in the $L'$ band, maintain single-mode behavior, provide achromatic splitting ratios, and achieve deep nulling ratios (raw $\sim 10^{-2}$ , self-calibrated $\sim 10^{-5}$ ) to suppress stellar leakage effectively.

2. Methodology

Device Design and Fabrication

Architecture: The device implements a double-Bracewell scheme, cascading three $2\times2$ directional couplers (DC1, DC2, DC3). This creates two nulled outputs that are centro-symmetric to each other, enabling self-calibration (kernel nulling) to cancel out residual phase errors.
Material: The beam combiner was fabricated in Gallium Lanthanum Sulfide (GLS), a chalcogenide glass transparent from 0.5 to 9 $\mu$ m with a refractive index of $\approx 2.31$ .
Technique: Ultrafast Laser Inscription (ULI) was used to write waveguides inside the GLS slab. The process utilized a Ti:Sapphire femtosecond laser (800 nm, <50 fs pulses) in a thermal fabrication regime.
Key Components:
- Directional Couplers: Asymmetric couplers were designed to achieve an achromatic splitting ratio across the $L'$ band.
- Y-Junctions: Used as photometric taps to monitor flux (initially 40/60 split, with a plan for 20/80 in future versions).
- S-bends: Cosine S-bends were implemented to route waveguides and minimize crosstalk and bending losses.

Experimental Testbed

Setup: A double-Michelson interferometer setup generated four coherent broadband beams simulating the four VLTI telescopes.
Sources: A supercontinuum source (2–5 $\mu$ m) filtered to the $L'$ band (3.65–3.85 $\mu$ m) and a blackbody source.
Control: Motorized delay lines (M2, M3, M4) controlled the optical path difference (OPD) to simulate fringe tracking and phase shifts.
Detection: An InSb camera recorded the output intensities at room temperature without active polarization control.

3. Key Contributions

First Broadband $L'$ 4T-Nuller: This work presents the first experimental measurement of a broadband deep null obtained with a four-telescope integrated optics beam combiner in the mid-infrared.
ULI in Chalcogenide Glass: It demonstrates the successful application of ULI in GLS glass for complex, multi-telescope interferometric circuits, proving the feasibility of mid-infrared photonic chips for astronomical instrumentation.
Self-Calibration Demonstration: The paper validates the double-Bracewell architecture's ability to produce self-calibrated nulls, which are robust against differential piston and phase errors.
Broad Null Reproduction: The team experimentally reproduced the $\theta^6$ broad null dependency (Angel & Woolf scheme), theoretically allowing for wider angular rejection of starlight.

4. Key Results

Optical Performance

Modal Behavior: The device operates in a single-mode regime with a Mode Field Diameter (MFD) of 26–28 $\mu$ m and high circularity (0.95).
Splitting Ratios:
- Central Coupler (DC3): Achieved a balanced, achromatic 50/50 splitting ratio across 3.65–3.85 $\mu$ m using a 7.5 mm interaction length.
- Side Couplers (DC1, DC2): Exhibited an achromatic but unbalanced 40/60 splitting ratio. Future designs will optimize interaction lengths (6.5 mm for side, 7.5 mm for center) to balance all couplers.
- Y-Junctions: Measured at 40/60; a redesigned geometry targeting 20/80 was validated for future iterations.
Throughput: The total measured throughput is 37% (including Fresnel losses). Theoretical calculations suggest this will rise to ~49% after applying anti-reflection coatings. Propagation losses were measured at $\approx 0.28$ dB/cm.
Polarization: Differential birefringence was quantified. While some elliptical polarization occurs, the differential birefringence between waveguides is low, with an average polarization angle mismatch of $6^\circ \pm 2^\circ$ .

Nulling Performance

Raw Null ( $r_n$ ): Measured at $8.13 \pm 0.03 \times 10^{-3}$ . This value is limited by the thermal background noise floor of the room-temperature setup and residual phase errors from the delay lines.
Self-Calibrated Null ( $s_c_n$ ): Measured at $1.14 \pm 0.01 \times 10^{-3}$ . This demonstrates the effectiveness of the kernel nulling technique in canceling common-mode errors.
Broad Null ( $\theta^6$ ): The team successfully simulated the Angel & Woolf telescope configuration. The central broad null depth was measured at $\sim 7.5 \times 10^{-3}$ , and the angular dependence fit a power law of $c = 6.0 \pm 0.1$ , confirming the $\theta^6$ behavior.

5. Significance and Future Outlook

Scientific Impact: This device is the core component of the NOTT instrument, which will be the first nulling instrument in the southern hemisphere. It enables the direct spectroscopy of exoplanets at the water-ice line, crucial for understanding planet formation and searching for biosignatures.
Technological Milestone: The successful fabrication and testing of a 4T-IO combiner in the mid-infrared validates the path toward miniaturized, stable, and high-performance interferometric instruments.
Next Steps:
- Cryogenic Testing: The device will be tested in cryogenic conditions to eliminate thermal background noise, which is currently the limiting factor for raw null depth.
- Coating: Application of anti-reflection coatings to boost throughput to the required ~50%.
- Polarization Control: Implementation of strategies (e.g., Wollaston prisms or waveplates) to mitigate polarization-induced null degradation.
- Post-Processing: Integration with the GRIP software pipeline to apply numerical self-calibration, aiming to reach the target self-calibrated null of $10^{-5}$ .

In conclusion, the paper confirms that the NOTT 4T-nuller is a viable, high-performance solution for mid-infrared nulling interferometry, meeting critical design requirements and paving the way for future exoplanet characterization at the VLTI.