🔭 astrophysics

Tricouplers for nulling interferometry with photonic integrated circuits

This paper investigates the potential of photonic integrated circuit tricouplers to achieve the precise, achromatic starlight suppression required for future space-based nulling interferometry missions, demonstrating their nulling capabilities and characterizing chromatic effects in laboratory settings to support the development of next-generation exoplanet observatories.

Original authors: Anusha Pai Asnodkar, Nemanja Jovanovic, Harry-Dean Kenchington Goldsmith, Ahmed Sanny, Yoo Jung Kim, Hani Nejadriahi, Isabelle Rivera, Michael P. Fitzgerald, Dimitri Mawet, Pradip Gatkine

Published 2026-07-30

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Anusha Pai Asnodkar, Nemanja Jovanovic, Harry-Dean Kenchington Goldsmith, Ahmed Sanny, Yoo Jung Kim, Hani Nejadriahi, Isabelle Rivera, Michael P. Fitzgerald, Dimitri Mawet, Pradip Gatkine

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 the night sky as a giant, glittering stage where stars are the blindingly bright spotlights and planets are the shy, tiny actors trying to whisper their lines from the shadows. For a long time, our telescopes have been like people trying to hear a mouse squeak while standing next to a jet engine; the star's glare is just too overwhelming to let us see the faint, rocky worlds or giant gas planets that might be hiding nearby. These hidden worlds are crucial because they hold the secrets to how planets form and whether they could ever support life. To solve this, astronomers are building a new kind of "super-cameras" called nulling interferometers. Think of these as a team of detectives who don't just look at the light; they actively cancel out the star's glare by mixing its light waves in a very specific way, like noise-canceling headphones for the universe. If they can get the waves to cancel perfectly, the star disappears, and the faint planet becomes visible. But building these machines with big, heavy mirrors and lenses is tricky because they are sensitive to bumps, temperature changes, and misalignment. This is where a new technology called "photonic integrated circuits" (PICs) comes in. Imagine shrinking a whole laboratory of mirrors and lenses down onto a tiny chip, the size of a fingernail, made of glass and silicon. These chips are robust, small, and can be mass-produced, offering a promising way to build the next generation of planet-hunting tools.

The paper you are about to read explores a specific, clever component on these tiny chips called a "tricoupler." You can think of a tricoupler as a three-lane highway for light. In a standard setup, you might have two lanes merging, but this device has three lanes that are perfectly symmetrical. The scientists designed it so that if you send light into the two outer lanes with a specific timing difference (a phase shift of $\pi$ ), the light waves crash into each other and cancel out completely in the middle lane, creating a "null" or a zone of darkness. This is the magic trick needed to hide the star. The researchers tested these chips in their lab using light at a wavelength of 1.55 $\mu$ m (which is in the infrared part of the spectrum, similar to the H-band used in astronomy). They wanted to see if these tiny, flat, planar tricouplers could actually create deep, dark nulls and if they could do it across a wide range of colors (broadband light), not just a single color.

The team found that these tricouplers are indeed very good at their job. In tests using a single color of light (monochromatic), they achieved a "null depth" of about -53.9 dB (which is a ratio of $4.1 \times 10^{-6}$ ) when they filtered out one type of light polarization. This means they successfully blocked out 99.9996% of the star's light in that specific lane. Even without the special filter, they reached depths around -28 to -32 dB. This is a significant achievement, suggesting that the basic design of the tricoupler works well and that the flat, planar shape of the chip doesn't ruin the cancellation effect for the star's light. They also discovered that adding a "crossing" where waveguides crisscross each other on the chip didn't significantly mess up the performance, which is great news because it means they can build more complex, crowded circuits on these small chips without losing performance.

However, the story isn't a perfect victory lap yet. When the scientists tried to use a wide range of colors (broadband light) to simulate real-world conditions, the "darkness" wasn't quite as deep, hovering around -30 dB. The paper suggests that the main culprit for this isn't the tricoupler itself, but rather the "thermo-optic phase modulators" (TOPMs) used to control the timing of the light. These modulators work by heating up a tiny part of the chip to change how fast light travels through it. The problem is that this heating effect changes differently depending on the color of the light. The authors measured that the voltage needed to create the perfect cancellation shifts as the color changes, causing the "null" to blur out across the spectrum. They showed that this chromatic effect is inherent to the heater technology they are using, not a flaw in the tricoupler design.

The researchers also ruled out a few other potential problems. They found that the tricouplers are surprisingly robust against "intensity imbalances," meaning that even if the light coming in from the two sides isn't perfectly equal in brightness, the device can still create a very deep null. They also confirmed that the waveguide crossings don't scatter light in a way that ruins the experiment. However, they did find that the devices are highly sensitive to polarization. The fact that the null depth jumped dramatically when they added a polarization filter suggests that the chip treats different "orientations" of light waves differently (birefringence), which limits how dark the null can get without that filter.

In conclusion, this paper demonstrates that photonic tricouplers are a viable and powerful tool for building compact, high-contrast instruments for finding exoplanets. The devices successfully create achromatic (color-independent) nulls for on-axis starlight by design, but their real-world performance is currently held back by the color-dependent nature of the heating elements used to control them. The authors suggest that future improvements will need to focus on creating "achromatic" phase shifters that work the same way for all colors, and perhaps designing circuits that handle different polarizations separately. While they haven't solved every problem, they have proven that the core idea works, paving the way for more complex, multiplexed networks of these chips that could one day help us see Earth-like worlds orbiting distant stars.

Technical Summary: Tricouplers for Nulling Interferometry with Photonic Integrated Circuits

Problem Statement
Conventional exoplanet detection techniques (transits, radial velocities, and direct imaging) struggle to observe Solar System analog gas giants and habitable-zone terrestrial planets. These objects are often located at orbital separations where transit probabilities are low, radial velocity signals are weak, and angular separations fall below the inner working angles of traditional coronagraphs. Long-baseline nulling interferometry offers a solution by suppressing starlight to detect faint planetary signals at high spatial resolution. However, achieving the necessary contrast requires exquisite control of beam combination and optical path length. Traditional bulk-optic architectures are increasingly sensitive to alignment errors, thermal drifts, and mechanical perturbations. While Photonic Integrated Circuits (PICs) offer a compact, scalable, and stable alternative leveraging telecommunications advancements, their current design and manufacturing precision remain insufficient for the stringent requirements of exoplanet instrumentation. Specifically, there is a need to characterize the nulling capabilities of PIC-based devices, particularly regarding their achromatic performance and sensitivity to chromatic effects introduced by active components.

Methodology
The authors investigated the nulling performance of lithographically fabricated photonic tricouplers on a silica-on-silicon platform. The devices consist of Ge-doped SiO2 waveguide cores with a 2% index contrast. The experimental setup involved:

Device Architecture: The study utilized single-input devices where integrated Y-splitters generated two coherent inputs on-chip to simulate two telescope apertures. The core components included passive splitters (Y-splitters, Wavelength Independent Couplers [WINCs], and 2x2 Multimode Interferometers [MMIs]) and active devices featuring 1x2 Mach-Zehnder Interferometers (MZIs) for intensity matching and tricouplers for beam combination.
Active Control: Thermo-optic phase modulators (TOPMs) were employed to induce phase shifts by locally heating the waveguides, altering the refractive index. Voltage control was managed via a multichannel controller, and the chip was mounted on a thermoelectric cooler (TEC) for temperature stabilization.
Experimental Modes: Data was acquired in two modes:
1. Monochromatic Photometry: Using a 1590 nm laser and a Xenics camera to measure all outputs simultaneously. Polarization filtering was applied using a polarizer and a pinhole mask to isolate the nulling port and minimize scattering contamination.
2. Broadband Spectral Monitoring: Using a superluminescent diode source and an Optical Spectrum Analyzer (OSA) to measure spectral performance across the H-band (~1550 nm).
Device Variants: Three specific tricoupler configurations were tested: a 1x3 tricoupler, a 1x5 tricoupler, and a 1x5 tricoupler incorporating a 76° waveguide crossing to evaluate the impact of crossings on null depth.

Key Contributions

Characterization of Passive Splitting: The authors evaluated the chromaticity of Y-splitters, WINCs, and 2x2 MMIs, identifying Y-splitters as providing the most achromatic 50:50 splitting across the bandpass.
Tricoupler Nulling Validation: The study experimentally validated planar tricouplers as beam combiners capable of producing achromatic nulls for on-axis starlight while allowing off-axis planet light to pass.
Analysis of Chromatic Limitations: The work quantified the chromatic degradation introduced by TOPMs. By comparing the wavelength-dependent null voltages of the tricoupler circuits against 1x2 MZIs, the authors isolated the intrinsic chromaticity of the phase shifter from that of the interferometric circuit.
Waveguide Crossing Impact: The inclusion of a waveguide crossing in the 1x5 device demonstrated that such crossings do not significantly degrade null depth, supporting the feasibility of compact, complex routing in future PICs.

Results

Monochromatic Performance: In monochromatic light (1590 nm), the devices achieved null depths ranging from -28.6 dB (1x3 device) to -32.0 dB (1x5 with crossing).
Polarization Effects: A significant improvement was observed when polarization filtering was applied to the 1x5 device with a crossing, achieving a null depth of -53.9 dB (4.1 × 10⁻⁶). This indicates that device birefringence, causing different null conditions for orthogonal polarization states, was a limiting factor in unfiltered measurements.
Broadband Performance: Polychromatic null depths were consistent with monochromatic measurements at approximately -30 dB. However, the optimal voltage required to achieve a null shifted with wavelength.
Chromaticity Analysis: The study found that the slope of the null voltage squared ( $dV^2/d\lambda$ ) increased with heater power. When normalized by heater power ( $V^2$ ), the slopes for both MZI and tricoupler devices collapsed to a near-constant value ( $1/\lambda$ ). This confirms that the observed chromatic degradation is primarily driven by the intrinsic wavelength dependence of the thermo-optic phase shifters rather than the tricoupler architecture itself.
Intensity Sensitivity: For the deepest polarization-filtered nulls, the null depth was highly sensitive to intensity imbalances, degrading by 1–2 orders of magnitude with a 0.2 V deviation from optimal settings.

Significance and Claims
The paper claims that integrated photonic tricouplers offer a promising platform for compact, broadband nulling interferometry, capable of achieving monochromatic null depths approaching -55 dB and polychromatic null depths $\lesssim$ -30 dB over a 200 nm bandpass. These results are comparable to, and in some cases exceed, recent performance of 3D beam combiners fabricated by ultrafast laser inscription.

The authors modestly conclude that while the tricoupler architecture is inherently achromatic, current broadband performance is limited by the chromatic response of the thermo-optic phase shifters and device birefringence. The work establishes a foundation for future high-contrast instruments, such as the Habitable Worlds Observatory, by demonstrating that waveguide crossings can be integrated without performance loss and by identifying the need for achromatic phase shifters and intensity modulators to further improve null depths. The single-input characterization presented serves as a precursor to future two-input interferometric operations that would enable simultaneous nulling and wavefront sensing.

More like this