On-sky demonstration of second-stage wavefront control with a photonic lantern

TL;DR

The paper demonstrates on-sky, closed-loop second-stage wavefront control using a 19-port photonic lantern at $1550\ \mathrm{nm}$ integrated with ShaneAO. By calibrating the PL response and operating with a leaky-integrator controller, the authors achieved a reduction of PL-measured RMS wavefront error from about $3.8\ \mathrm{nm}$ to $1.3\ \mathrm{nm}$ and a measurable improvement in the final PSF, validating focal-plane wavefront sensing as a minimally invasive retrofit. Key contributions include the successful on-sky deployment of an undispersed PL for quasi-static NCPA correction, a practical two-stage AO workflow, and a discussion of throughput and alignment bottlenecks that presently limit performance. The work points to practical pathways for retrofitting existing AO systems with PL-based sensing to enhance high-contrast imaging without requiring extensive new instrumentation.

Abstract

Ground-based direct imaging of exoplanets at high contrast requires precise correction of atmospheric turbulence using adaptive optics (AO). The planet-to-star contrast ratio at small angular separations from the host star is often limited by non-common-path aberrations (NCPAs) seen only in the science plane. The photonic lantern (PL) can be used to sense aberrations at the final science imaging plane. This enables a two-stage wavefront control architecture, in which the first-stage wavefront sensor senses atmospheric turbulence and the PL senses NCPAs and other aberrations not seen by the first stage. We demonstrate closed-loop control of residual wavefront errors using a non-dispersed PL after first-stage AO correction on the Shane 3m telescope at Lick Observatory. Our results show that non-dispersed PLs can be used for second-stage wavefront sensing, enabling performance improvements via minimally invasive retrofits to existing AO systems.

Figures (13)

• Figure 1: The experimental setup. The middle part of the diagram (in the blue box) leading from the ShaneAO DMs describes the optical setup installed for this work, and the lower part (in the red box) describes the software pipeline that was used. A representative PL image from this setup is shown in the top right.
• Figure 2: The optical layout of the PLIU and output imaging configuration used in this experiment. The dotted line from the beam splitter to the injection-monitoring camera is a back-reflection from the PL input.
• Figure 3: The PLIU installed within ShaneAO. Light coming into the PLIU is converging and comes to a focus on a mirror, diverges as it passes through a beam splitter, and is collimated and focused along three separate paths.
• Figure 4: Outputs from the two aligning cameras at the best alignment. We expect to see a faint ring with no light in its center on the injection-monitoring camera when well aligned.
• Figure 5: Latency of the control loop as measured on the control computer. Three measurements were taken above 1000 ms; for clarity, these are not shown.