26th Australian Institute of Physics Congress and the Australian and New Zealand Conference on Optics and Photonics

Hybrid Mode-Selective Photonic Lanterns for In-Situ Wavefront Sensing and Optimal Injection into Photonic Chips at VLTI

Not scheduled

20m

Belinda Hutchinson Building (The University of Sydney )

Contributed Oral ANZOS | Photonics and Optics (ANZCOP)

Speaker

Daniel Dahl (The University of Sydney)

Description

Hybrid mode-selective photonic lanterns (HMS-PLs) are an emerging astrophotonic technology that enable simultaneous mode-selective light injection and wavefront sensing. Unlike conventional photonic lanterns, HMS-PLs selectively couple the fundamental LP01 mode from a multimode input into a dedicated single-mode output, while higher-order modes are distributed among the remaining outputs for wavefront sensing. This makes HMS-PLs attractive for astronomical instrumentation, where efficient coupling of diffraction-limited light and low-order wavefront sensing are critical.

We present the characterisation of prototype six-core HMS-PLs being developed for Seidr, part of the Asgard Suite for the Very Large Telescope Interferometer. A key challenge for deploying HMS-PLs in closed-loop adaptive optics systems is determining the lantern transformation between the multimode input and single-mode outputs, particularly in situ at a telescope where direct field measurements are impractical.

To address this, we investigate complementary characterisation techniques. Off-axis digital holography is used as a laboratory validation tool to reconstruct the complex modal fields emerging from the lantern outputs and verify device performance. In contrast, high-dimensional Stokes measurements and machine-learning approaches are explored as practical in-situ methods that recover the lantern transformation using only output intensity measurements. The Stokes approach enables reconstruction of the lantern transmission matrix, while neural networks learn the transformation directly from measured single-mode powers.

Using these methods, we investigate wavefront correction strategies based solely on HMS-PL output powers. Several neural network architectures are compared against one another and against transmission-matrix-based reconstruction from the high-dimensional Stokes method for estimating the incident wavefront and generating correction signals for an upstream deformable mirror. These results demonstrate the potential of HMS-PLs as compact, integrated devices for simultaneous wavefront sensing and mode-selective injection in future high-contrast astronomical instrumentation.