🔬 applied physics

Technologies and novel components for broadband splitting and coupling in pairwise and nulling interferometry

This paper presents the design and characterization of broadband photonic components, including tapered tri-couplers, directional couplers, and an achromatic intensity modulator on silicon nitride and silicon oxide platforms, which enable high-performance pairwise and nulling interferometry for exoplanet detection across the J- and H-bands.

Original authors: Harry-Dean Kenchington Goldsmith, Nemanja Jovanovic, Anusha Pai Asnodkar, Sanny Ahmed, Elsa Huby, Sylvestre Lacour, Michael Fitzgerald, Yoo Jung Kim, Pierre Labeye, Nicolas Dunoyer, Michael Ireland, S

Published 2026-07-28

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Harry-Dean Kenchington Goldsmith, Nemanja Jovanovic, Anusha Pai Asnodkar, Sanny Ahmed, Elsa Huby, Sylvestre Lacour, Michael Fitzgerald, Yoo Jung Kim, Pierre Labeye, Nicolas Dunoyer, Michael Ireland, Stephen Madden

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

The Cosmic Game of Hide-and-Seek

Imagine trying to hear a tiny, whispering firefly buzzing next to a blindingly bright stadium floodlight. That is the challenge astronomers face when trying to find planets orbiting other stars. The star is the floodlight, and the planet is the firefly. Because the star is so incredibly bright, its light usually washes out the faint glow of the planet, making it impossible to see. To solve this, scientists use a clever trick called "interferometry." Think of it like having two pairs of eyes looking at the same scene. If you line them up just right, you can make the light from the bright star cancel itself out—like noise-canceling headphones for light—while the light from the planet, which is coming from a slightly different angle, stays visible.

To make this work, scientists need to split the light coming from the telescope and then recombine it with extreme precision. In the past, they used bulky mirrors and lenses, but these are heavy and hard to keep perfectly aligned. The new frontier is "photonic integrated circuits" (PICs). Imagine shrinking an entire optical laboratory down to the size of a computer chip. Instead of big mirrors, light travels through tiny channels carved into glass or silicon. The goal is to build these chips so they can handle a wide range of colors (broadband) of light without getting confused or losing the signal. If we can master this, we might finally be able to spot Earth-like worlds around distant suns and check if they have the ingredients for life.

The Paper's Story: Tuning the Cosmic Tuning Forks

This paper is all about building better, more versatile "tuning forks" for these light chips. The researchers, a team from universities and labs across Australia, the US, and France, are designing new components that can split and mix light across a broad spectrum of colors, specifically for the near-infrared part of the light spectrum (the J-band and H-band). They aren't just tweaking old designs; they are inventing new shapes for the light channels to make them work better.

The team focused on two main types of "chips": one made of silicon nitride (a hard, glass-like material) and another made of silicon oxide (essentially high-tech glass). They simulated how light behaves in these materials to design two key inventions: tapered tri-couplers and tapered directional couplers.

Think of a standard light splitter like a Y-junction, where a single road splits into two. It works, but it's not very good at handling different colors of light at once, and it often loses some of the light in the process. The researchers proposed a "tapered tri-coupler" instead. Imagine a three-lane highway where the two outer lanes slowly widen or narrow as they merge with the middle lane. By carefully shaping these lanes (tapering them), the light can be split perfectly between the two outer lanes while keeping almost none of it stuck in the middle lane. In their simulations for the silicon nitride chips, these new splitters lost less than 1% of the light across the entire J-band (from 0.95 to 1.8 micrometers). That's a huge improvement over older designs.

For the silicon oxide chips, which are used for the H-band (1.5 to 1.8 micrometers), the team designed a similar tapered tri-coupler but with a specific superpower: nulling. This is the "noise-canceling" part. When they send light from a star into the two outer lanes with a specific timing offset (180 degrees out of phase), the light cancels itself out in the middle lane, creating a "null." This is how you hide the star. However, they also need to make sure the light from a planet (which arrives with a different timing) doesn't get cancelled out. Their simulations showed that their optimized design could hide the starlight while letting more than 97.8% of the planet's light through across the H-band. That's a massive win for spotting those faint fireflies.

The paper also introduces a brand-new gadget called a Chromatically Controlled Achromatic Intensity Modulator (CCAIM). This is a bit like a smart dimmer switch for light. In a nulling interferometer, you need to balance the intensity of the light beams perfectly to get the best cancellation. Usually, this is hard to do because different colors of light behave differently. The CCAIM uses a heater to change the speed of light in one part of the chip, but it's paired with a special tapered coupler that compensates for the color differences. The simulations suggest this device can actively adjust the brightness of the light beams to be achromatic (working the same for all colors) across a range of 150 to 300 nanometers. While this is currently just a simulation, it suggests a way to actively tune the system to get deeper "nulls" (better star suppression) without needing complex external equipment.

The researchers are also looking ahead to the mid-infrared, a region of light where planets glow even brighter. They suggest that while silicon chips stop working well at these longer wavelengths, a material called chalcogenide glass (a mix of sulfur, selenium, or tellurium) could take over. They propose a roadmap to build similar tapered couplers in this new material, which could eventually help space telescopes like the Habitable Worlds Observatory or the LIFE project detect oxygen or ozone on distant Earth-like planets.

However, the paper is careful to note what is and isn't proven yet. The silicon nitride results are based on simulations, though some preliminary measurements on the tapered tri-couplers by colleagues in France showed promising results with negligible excess loss. The silicon oxide nulling results are also simulations, and the authors warn that real-world manufacturing errors (like a waveguide being slightly wider or narrower than planned) could affect performance. For instance, a tiny shift in the width of the waveguide could increase the loss of planet light from under 2.2% to 4.5% at the long-wavelength end. The CCAIM is purely a proof-of-concept simulation; the authors explicitly state that further work is needed to see if it can maintain its achromatic performance across all possible settings.

In short, this paper doesn't claim to have built the final machine that will find alien life tomorrow. Instead, it provides the blueprints and the "aha!" moments for the next generation of components. It shows that by changing the shape of the light channels from simple straight lines to carefully tapered curves, we can build chips that are more efficient, handle more colors of light, and are better at hiding stars to reveal planets. It's a step toward making the complex, fragile art of astronomical interferometry into something robust, scalable, and ready for the next giant leap in space exploration.

Technical Summary: Technologies and Novel Components for Broadband Splitting and Coupling in Pairwise and Nulling Interferometry

Problem Statement
The continued development of astronomical photonic integrated circuits (PICs) for ground- and space-based interferometry is hindered by the challenge of achieving broadband achromatic performance. While passive and active photonic components are central to integrating adaptive optics, beam combination, and spectroscopy onto single chips, standard components often suffer from excess loss or narrow operational bandwidths. Specifically, there is a need for improved beam splitters and combiners that function effectively across the astronomical J-band (0.95–1.35 µm) and H-band (1.5–1.8 µm) to support both pairwise interferometry and nulling interferometry (used for exoplanet detection). Furthermore, as the field moves toward the mid-infrared (MIR) for detecting Earth-like exoplanets, new material platforms and component designs are required to overcome the absorptive limitations of silicon at longer wavelengths.

Methodology
The authors investigated and simulated novel photonic components using two distinct material platforms: Silicon Nitride (Si $_3$ N $_4$ ) for the J-band and Germanium-doped Silicon Oxide (SiO $_2$ ) for the H-band. The study utilized RSoft BeamPROP simulations (considering TE polarization) to design and optimize custom-tapered components, moving beyond standard Y-junctions and conventional directional couplers.

Silicon Nitride (J-band): Using STMicroelectronics technology, the team designed tapered tri-couplers and tapered directional couplers. The optimization focused on varying the coupling-region length and the width deviation ( $\Delta W$ ) of the outer waveguides relative to a central waveguide to achieve adiabatic, broadband splitting.
Silicon Oxide (H-band): Using Enablence technology with a low-confinement waveguide (3.4 $\times$ 3.4 $\mu$ m), the team simulated a two-dimensional tapered tri-coupler for nulling interferometry. The design aimed to facilitate destructive interference for on-axis starlight (180° phase offset) while maximizing throughput for off-axis exoplanet light (0° phase offset).
Active Control: A Chromatically Controlled Achromatic Intensity Modulator (CCAIM) was proposed. This device combines a Y-junction, a thermo-optic phase shifter, and a tapered directional coupler to actively balance optical intensity between interferometer arms.
Future Roadmap: The paper outlines a transition to Chalcogenide (ChG) glass platforms for the MIR (L- and N-bands), proposing the adaptation of the tapered components developed for Si $_3$ N $_4$ and SiO $_2$ to these new materials.

Key Contributions and Results

Tapered Tri-couplers (Si $_3$ N $_4$ ): Optimized tapered tri-couplers achieved less than 1% excess loss across the entire J-band. These devices successfully split light equally between two outer waveguides while minimizing power in the central output port, serving as a viable, low-loss alternative to Y-junctions and MMIs. Preliminary measurements by LETI confirmed negligible excess loss, though independent verification of the central port power is pending.
Tapered Directional Couplers (Si $_3$ N $_4$ ): Simulations demonstrated that tapered directional couplers can provide broadband 50:50 and 40:60 splitting ratios suitable for beam combination. The study found that waveguide widths further from the global width (600 nm) produced flatter broadband responses, with an 800 nm input width maintaining a 40:60 split across the full bandwidth.
Tapered Tri-couplers (SiO $_2$ ): For the H-band nulling interferometer, optimized tapered tri-couplers achieved less than 2.2% exoplanet throughput loss across the band. The study identified a trade-off between maximizing exoplanet throughput and maintaining phase-sensing sensitivity (fringe tracking). Specific parameter sets were identified that avoid degeneracies in phase sensing, allowing for active null control.
Chromatically Controlled Achromatic Intensity Modulator (CCAIM): The authors introduced a new device concept that uses a tapered directional coupler to compensate for the chromaticity of a thermo-optic phase shifter. Simulations showed that a single heating source could achieve programmable attenuation (splitting ratios from 100:0 to 35:65) with achromatic behavior over a 150–200 nm bandwidth. The device demonstrated a simulated extinction ratio exceeding 30 dB across a 200 nm waveband.
Mid-Infrared Roadmap: The paper proposes a roadmap for extending these low-loss, broadband concepts to Chalcogenide platforms for the L- and N-bands. It highlights the potential of Selenium-doped ChG (Se-ChG) for the N-band to avoid sulfur absorption features, aiming to support future space missions like the Habitable Worlds Observatory (HWO) and the Large Interferometer For Exoplanets (LIFE).

Significance and Claims
The paper claims that these developments illustrate the "growing maturity" of broadband photonic components for astronomical interferometry. The primary significance lies in the demonstration that custom-tapered components can overcome the bandwidth and loss limitations of standard PIC elements in both low-confinement (SiO $_2$ ) and high-confinement (Si $_3$ N $_4$ ) platforms.

The authors modestly state that while the CCAIM and tapered couplers show promise, further experimental validation is required to confirm fabrication tolerances, modal analysis for wider waveguides, and the full extent of achromatic performance. The work does not claim to have solved all challenges for MIR interferometry but establishes a foundational pathway using tapered tri-couplers and directional couplers to replace conventional MMIs and Y-junctions. The ultimate goal is to enable fully integrated, low-loss, and scalable photonic interferometers capable of detecting Earth-like exoplanets across the near- and mid-infrared spectrum.

The Cosmic Game of Hide-and-Seek

The Paper's Story: Tuning the Cosmic Tuning Forks

More like this