News from Photonics
Actuality from the project
WP1 : Visible FIRST upgrade @ LIRA
An injection bench designed to coherently simulate the injection of five beams has been installed and aligned. It is powered by a supercontinuum source, which allows either a visible beam to be emitted for the characterisation of FIRST components, or a beam in the near-infrared (see WP2). The bench is therefore ready to carry out photometric and interferometric characterisation of integrated optical chips.
A 5T-2D chip was designed by G. Martin and manufactured by TEEM Photonics using Ag ion-exchange technology. Compared with the previously used K-ion-exchange technology, this allows for a greater refractive index contrast between the core and the cladding of the waveguides, and therefore better beam confinement. Consequently, the bends can be tighter and the chip is more compact, minimising losses
A 5T-3D complementary chip was fabricated using laser photolithography at the Hubert Curien Laboratory in Saint-Étienne. This technology enables the creation of waveguides in three dimensions, without being restricted to propagation in a single plane, thereby avoiding waveguide crossings and cross-talk effects. Characterisation of the chip showed that it was not single-mode at visible wavelengths, but was single-mode in the near-infrared. This chip will therefore be characterised for a potential application under WP2, in the near-infrared, as part of the PLANETES project.
Alongside these developments, another avenue in photonics is being explored within the FIRST-PL project: this involves the use of a photonic lantern (a multimode optical fibre on one end, which converts the beam into 19 beams across as many single-mode outputs, corresponding to coupling into a specific spatial mode). The main advantage of this component is its high photon transmission efficiency, thanks to improved coupling due to multimodality and an adiabatic transition to single-mode fibres. We have developed a strategy for observations and data analysis, which has enabled us to validate the commissioning of this mode on the Subaru telescope, with applications in spectro-astrometry (Kim et al. 2025), H-alpha differential imaging and spectral imaging.
WP2 : J-band referencing @ LIRA
For the characterisation of components in the near-IR, the injection bench mentioned in WP1 has been supplemented with a spectrometer optimised for the near-IR. The aim of this WP is to develop a high-performance chip to extend the VLTI into the near-IR wavelength range, which is the focus of the PLANETES project (ERC, PI S. Lacour). Several technological solutions are under consideration and will be compared on this test bench:
- A chip designed and manufactured by VLC Photonics has been characterised in terms of transmission: approximately 20 per cent, which is still below the target (>50 per cent). Interferometric characterisation will be carried out in 2026.
- a 5T silicon nitride chip, optimised by P. Labeye as part of a collaboration with CEA Leti, funded by the PEPR, and manufactured by ST Microelectronics. Preliminary characterisation shows that the losses induced by the individual functions (coupler, cross, turn) are acceptable. Characterisation will continue on the test bench at Meudon once the inputs have been fibre-connected.
- A chip manufactured by TeemPhotonics using silver ion diffusion technology is currently under development; preliminary characterisation indicates a transmission rate of over 45 per cent, making it highly promising.WP3 : H-band kernel-nulling @ Lagrange
The appointment of Marc-Antoine Martinod on 1 April 2025 has helped to speed up the upgrade work on the PHOTONICS bench, which now includes the option to connect a wavelength-tunable fibre laser source and an optical spectrum analyser, and benefits from a software control environment developed by Vincent Foriel (a PhD student on the project).
Meticulous, long-term work carried out by MA. Martinod has made it possible to identify and eliminate the main source of cross-talk observed in the optical signals at the output of our SiN component, thereby enabling us to improve the nulller’s monochromatic rejection performance.
V. Foriel was thus able to put into practice the component optimisation algorithms developed in simulation, by controlling the thermo-optical actuators integrated into the SiN component of a ‘4×4’ (4-input, 4-output) nulling kernel consisting of a single MMI cavity. The experimental results are currently being compiled for a peer-reviewed article.
V. Foriel has also continued the work on formalisation required for the correct interpretation of high-contrast interferometric signals produced by the outputs of a nulling kernel, leading to the evaluation and comparison of several statistical detection tests. These results are also currently being consolidated in a peer-reviewed article.
MA. Martinod was able to demonstrate, using the same experimental setup, that the same 4×4 nulling kernel can also be used effectively as an ABCD recombiner. These experimental results will be presented at the SPIE 2026 conference.
F. Martinache took the opportunity of the WITSO workshop held at ESA to present the current status of ideas concerning the design of a flat space telescope (in fact, an imaging interferometer made up of photonic elements), capable of reconstructing high-resolution images.
As part of a collaboration with the University of Sydney, our group is making progress on the design of a kernel-nuller recombiner, which would be powered by four hybrid photonic lanterns and is intended to be integrated as a new module within the VLT/ASGARD instrument, known as SEIDR.
WP4 : L-band nuller @ IPAG
Pierre Labeye (CEA-Leti) designed the first generation of interferometric photonic circuits using SiGe technology. These were fabricated and characterised by the Leti teams. The photometric performance of the circuits exceeded expectations in terms of transmission and met the specifications for routing properties, which is a source of great satisfaction. The IPAG took delivery of the photonic circuits in December 2025, in line with the original schedule. We included circuits with active phase modulation in this fabrication run – the first of their kind in the mid-infrared – and their characterisation is currently underway.
Louis Gemmerlé began his PhD at IPAG in October 2025 and initially focused on developing a two-channel interferometric bench – a precursor to the MILENE bench – designed for the detailed characterisation of mid-infrared photonic circuits, with the assistance of Manon Lallement (postdoctoral researcher). MILENE-2T achieved first light in November 2025 and enabled the first SiGe circuit to be tested in interferometric mode. This allowed laser contrasts of 98 per cent to be measured, which is very promising for a first trial and in the absence of active control over MILENE’s properties. The expansion of MILENE to four channels has been initiated but requires significant effort in terms of control software, with support from the IPAG’s ELI and P2I departments.
Manon Lallement (awarded the Olivier Chesneau Prize in 2025 for the best thesis in high-angular-resolution astrophysics) worked with Louis Gemmerlé to characterise the spectral performance of SiGe photonic circuits with an AWG structure, which enable spectrometers to be integrated into optical chips. She was thus able to compare the results with simulations, which revealed discrepancies that have not yet been understood. A formal collaboration agreement between Leti and UGA has been drawn up to enable Manon to work at Leti.
At the same time, G. Martin has designed a four-channel ‘Double Bracewell’ interferometric circuit using lithium niobate photonic technology, which will be characterised during 2026.
Focus
With the characterisation benches now in place at the project’s three main partner laboratories, and an initial batch of components currently being characterised, all the pieces are in place for a wealth of experimental results, many of which will be presented at the SPIE conference scheduled to take place in Copenhagen in July 2026.
The characterisation of first-generation components enables each partner to assess the advantages and limitations of the photonic platforms under investigation. The project aims to achieve specific performance metrics (transmission, stability of phase-closing measurements, null depth, bandwidth, ability to control the electric field), and verifying these will require several fabrication cycles.
At Lagrange, for example, work to identify sources of cross-talk (mutual interference between optical signals) has helped to guide the launch of production of a new generation incorporating the lessons learnt: spacing the waveguides further apart, increasing the bending radii of the bends, and implementing solutions to eliminate evanescent coupling and facilitate the experimental measurement of high rejection by the nullifier. The design of this new component is currently underway in collaboration with Bright Photonics and is expected to be delivered to the laboratory before the end of 2026.
Relation with others WP
The project is linked to COMPACT-SPECTROGRAPHS, which explores various approaches to implementing single-mode spectroscopy. A key aspect of the work we are carrying out here is the breadth of the wavelength range covered by our components and the interferometric recombiners. Single-mode spectroscopy solutions can be elegantly integrated into interferometric recombiners for the purpose of spectral characterisation (as with the AWGs integrated into SiGe chips developed by CEA Leti and characterised at IPAG) or as a preliminary step to pre-compensate for the chromatic behaviour of our recombiners.
The latest technical developments planned as part of the project are expected to capitalise on this approach to functional hybridisation and lead to the realisation of integrated spectro-interferometers.
Relation with industries
The technical aspects of the project are systematically developed through collaboration between a research laboratory (driven by applications in astrophysics) and an industrial partner with expertise in a manufacturing process. The project’s partner companies are:
Publication links to the project
https://ui.adsabs.harvard.edu/abs/2025ApJ…993L…3K/abstract
https://ui.adsabs.harvard.edu/abs/2025A%26A…694A.277L/abstract
https://ui.adsabs.harvard.edu/abs/2024A%26A…691A.140V/abstract
https://ui.adsabs.harvard.edu/abs/2024Natur.634.1070X/abstract
https://ui.adsabs.harvard.edu/abs/2024SPIE13096E..0PV/abstract
https://hal.science/hal-05390426
codes développés et mis en ligne sur des dépôts publics :