Nulling at short wavelengths: theoretical performance constraints and a demonstration of faint companion detection inside the diffraction limit with a rotating-baseline interferometer

TL;DR

The paper presents the design, performance, and a demonstration of the Palomar Fiber Nuller, a rotating-baseline interferometer capable of detecting faint companions within the diffraction limit using nulling techniques.

Contribution

It introduces the final design and performance of the PFN and demonstrates faint companion detection through nulling-baseline rotation, advancing high-precision nulling interferometry.

Findings

Successful detection of a faint companion in eta Peg using nulling rotation

Achieved null-depth accuracy of a few times 10^-4 in NIR observations

Demonstrated potential for small-angle nulling with larger telescopes

Abstract

The Palomar Fiber Nuller (PFN) is a rotating-baseline nulling interferometer that enables high-accuracy near-infrared (NIR) nulling observations with full azimuth coverage. To achieve NIR null-depth accuracies of several x 10-4, the PFN uses a common-mode optical system to provide a high degree of symmetry, single-mode-fiber beam combination to reduce sensitivity to pointing and wavefront errors, extreme adaptive optics to stabilize the fiber coupling and the cross-aperture fringe phase, rapid signal calibration and camera readout to minimize temporal effects, and a statistical null-depth fluctuation analysis to relax the phase stabilization requirement. Here we describe the PFN final design and performance, and provide a demonstration of faint-companion detection by means of nulling-baseline rotation, as originally envisioned for space-based nulling interferometry. Specifically, the Ks-band null-depth rotation curve measured on the spectroscopic binary eta Peg reflects both a secondary star 1.08 +/- 0.06 x 10-2 as bright as the primary, and a null-depth contribution of 4.8 +/- 1.6 x 10-4 due to the size of the primary star. With a 30 mas separation at the time, eta Peg B was well inside both the telescope diffraction-limited beam diameter (88 mas) and typical coronagraphic inner working angles. Finally, we discuss potential improvements that can enable a number of small-angle nulling observations on larger telescopes.