China's Tiny Photonic Chip Spectrometer Achieves Major Astronomy Milestone

Chip-Scale Spectrometer Delivers Telescope-Grade Performance in 500 cm³

A research team at the Chinese Academy of Sciences (CAS) has successfully demonstrated a chip-scale integrated photonic spectrometer capable of performing high-resolution astronomical observations — a feat previously requiring instruments the size of a room. The device, which fits within a volume of just 500 cm³ (roughly the size of a coffee mug), achieved a spectral resolution exceeding 25,000 and a working bandwidth of more than 180 nm, marking what the team calls the first-ever observation of near-infrared solar Fraunhofer lines using a high-resolution integrated photonic spectrometer.

The findings, published in the peer-reviewed journal Photonics Research (DOI: 10.1364/PRJ.582324), represent a significant step toward miniaturizing the bulky optical instruments that have long dominated ground-based and space-based astronomy. The work was carried out by the astronomical photonics team at CAS's Nanjing Institute of Astronomical Optics & Technology (NIAOT).

Key Takeaways at a Glance

• Spectral resolution exceeds 25,000 — peaking at approximately 26,500 near the spectral center
• Working bandwidth spans over 180 nm in the near-infrared range
• Total system volume is under 500 cm³, more than 1,000× smaller than conventional meter-scale solar spectrometers
• Chip dimensions measure just 9.6 mm × 3.2 mm, fabricated on a low-loss silicon nitride (Si₃N₄) platform
• First demonstration of near-infrared solar Fraunhofer line observation using an integrated photonic spectrometer
• Hybrid dispersion architecture combines cascaded phase-modulated waveguide array chips with a cross-dispersion module

How the Technology Works: A Hybrid Dispersion Approach

Traditional astronomical spectrometers rely on large diffraction gratings or echelle gratings housed inside instruments that can stretch several meters in length. These systems deliver excellent performance but are expensive, heavy, and extremely difficult to deploy on small satellites or in remote observatory settings.

The NIAOT team took a fundamentally different approach. They designed a hybrid dispersion integrated photonic spectrometer that combines two key innovations: a cascaded phase-modulated waveguide array chip and an orthogonal (cross) dispersion module. The waveguide array chip handles primary spectral dispersion on the photonic circuit level, while the cross-dispersion module separates overlapping spectral orders — a technique conceptually similar to how echelle spectrometers work, but executed at a dramatically smaller scale.

The photonic chip itself was fabricated using a low-loss silicon nitride platform, a material widely used in integrated photonics for its excellent transparency across visible and near-infrared wavelengths and its compatibility with standard semiconductor manufacturing processes. The final chip measures a mere 9.6 mm × 3.2 mm — small enough to sit comfortably on a fingertip.

Performance That Rivals Room-Sized Instruments

The experimental results are striking. Across its full working range, the integrated spectrometer achieved a resolving power consistently above 20,000. Near the center of its spectral range, performance peaked at approximately 26,500 — a figure that places it in the same league as many conventional benchtop and even some facility-class spectrometers.

The system's total spectral coverage spans roughly 180 nm, which is broad enough to capture multiple scientifically important spectral features in a single observation. To put this in perspective, many existing integrated photonic spectrometers operate with bandwidths of only a few nanometers, making them unsuitable for broadband astronomical science.

Perhaps most impressively, the entire optical assembly — including the chip, cross-dispersion optics, and detector coupling — occupies less than 500 cm³. Traditional solar spectrometers that achieve comparable resolving power typically occupy volumes measured in cubic meters. The NIAOT system therefore represents a volume reduction of more than 3 orders of magnitude, or roughly a factor of 1,000.

First-Ever Fraunhofer Line Observation With an Integrated Photonic Spectrometer

The team validated their instrument by pointing it at the Sun and successfully resolving near-infrared Fraunhofer lines — the dark absorption lines in the solar spectrum caused by chemical elements in the Sun's outer atmosphere. These lines have been studied since the early 19th century and serve as critical benchmarks for spectrometer performance.

Observing Fraunhofer lines requires both high spectral resolution (to distinguish closely spaced lines) and sufficient bandwidth (to capture multiple lines across a wide wavelength range). The fact that a chip-scale device accomplished this task for the first time underscores the maturity of the underlying photonic technology.

This demonstration is particularly significant because it moves integrated photonic spectrometers from the laboratory proof-of-concept stage into the realm of real astronomical science. Previous integrated photonic spectrometer prototypes have generally been limited to narrow-band demonstrations or low-resolution measurements, making them impractical for most astrophysical applications.

Why This Matters for the Future of Astronomy

The miniaturization of astronomical instruments has been a long-standing goal in the field, driven by several converging needs:

• Space missions demand instruments that are lightweight, compact, and power-efficient
• Multi-object spectroscopy on large telescopes requires hundreds or thousands of spectrometer channels operating simultaneously
• Remote and autonomous observatories benefit from instruments with no moving parts and minimal maintenance requirements
• CubeSat and SmallSat platforms impose strict volume and mass constraints that conventional spectrometers cannot meet
• Cost reduction is essential as the astronomy community pursues increasingly ambitious survey programs

Integrated photonic spectrometers address all of these challenges simultaneously. By replacing bulk optical components with waveguide circuits etched onto a chip, researchers can achieve dramatic reductions in size, weight, and power (SWaP) — the three metrics that most constrain instrument design for space and remote applications.

How This Fits Into the Broader Photonics Landscape

The NIAOT work builds on more than a decade of international research into astrophotonics, a field that applies integrated photonics technologies to astronomical instrumentation. Groups at institutions including the University of Sydney, Leibniz Institute for Astrophysics Potsdam (AIP), and Heriot-Watt University in Scotland have all contributed pioneering work in this space.

However, most previous demonstrations have faced a fundamental trade-off between resolution and bandwidth. Arrayed waveguide gratings (AWGs), for instance, can achieve high resolution but typically cover only a narrow spectral window. Fourier transform spectrometers on chip can achieve broad bandwidth but often sacrifice resolution.

The NIAOT team's hybrid approach — combining a waveguide array with cross-dispersion optics — effectively breaks this trade-off. The result is a system that simultaneously delivers:

• High resolution (R > 25,000)
• Broad bandwidth (> 180 nm)
• Compact form factor (< 500 cm³)
• Compatibility with standard silicon nitride fabrication

This combination of attributes has not been previously demonstrated in any integrated photonic spectrometer, making the work a notable advance in the field.

Looking Ahead: Broader Bandwidth and Deep-Sky Targets

The research team has outlined several pathways for improving performance in future iterations. By optimizing the imaging optical system and adopting larger-format detector arrays, the working bandwidth could be expanded significantly beyond the current 180 nm — potentially covering the entire near-infrared astronomical J-band or H-band in a single device.

Additional improvements in waveguide loss and chip design could push the resolving power even higher, potentially reaching R > 50,000 — a threshold that would open the door to precision radial velocity measurements for exoplanet detection.

The team also envisions deploying arrays of such chip-scale spectrometers on the focal planes of large telescopes, enabling massively parallel multi-object spectroscopy. In this scenario, hundreds or thousands of miniature spectrometers could simultaneously analyze light from different celestial objects — a capability that would dramatically accelerate large-scale astronomical surveys.

For the broader photonics and AI-driven instrumentation community, this work highlights how advances in semiconductor fabrication and photonic circuit design are converging to create a new generation of scientific instruments. As machine learning techniques increasingly play a role in spectral analysis, calibration, and data reduction, the combination of compact photonic hardware and intelligent software could redefine how astronomical observations are conducted — both on the ground and in space.

The research was supported by CAS and published in the April 2025 issue of Photonics Research.