Worlds Next Door: A Candidate Giant Planet Imaged in the Habitable Zone of α Centauri A. II. Binary Star Modeling, Planet and Exozodi Search, and Sensitivity Analysis

The James Webb Space Telescope (JWST) observed our closest solar twin, α Centauri A (α Cen A), with the Mid-Infrared Instrument in the F1550C (15.5 μm) coronagraphic imaging mode at three distinct epochs between 2024 August and 2025 April. For the first time with JWST, we demonstrate the application of reference star differential imaging to simultaneously subtract the coronagraphic image of a primary star (α Cen A) and the point-spread function (PSF) of its binary companion (α Cen B) to conduct a deep search for exoplanets and exozodiacal dust emission. We achieve a typical 5σ point-source contrast sensitivity between ∼10⁻⁵ and 10⁻⁴ at separations ≳ 1″ and an exozodiacal disk (coplanar with α Cen AB) sensitivity of ∼5-8× the solar system’s zodiacal cloud around α Cen A. The latter is an extraordinary limit, representing the deepest sensitivity to exozodiacal disks achieved for any stellar system to date. Additionally, postprocessing with the principal-component-analysis-based Karhunen-Loéve image processing algorithm reveals a point source, called S1, in 2024 August, detected at signal-to-noise ratio of 4-6 (3.3-4.3σ), a projected separation of ≈1 . ″ 5 (2 au), and with an F1550C flux density (contrast) of ≈3.5 mJy (≈5.5 × 10⁻⁵). Various tests conducted with the available data show that S1 is unlikely to be a detector artifact or PSF-subtraction artifact and confirm that it is neither a background nor a foreground object. S1 is not redetected in two follow-up observations (2025 February and April). If S1 is astrophysical in nature, the only explanation is that it has moved to a region of poor sensitivity due to orbital motion. We perform PSF injection-recovery tests and provide 2D sensitivity maps for each epoch to enable orbital completeness calculations. Additional observations, with JWST or upcoming facilities, are necessary to redetect candidate S1 and confirm its nature as a planet orbiting our nearest solar-type neighbor, α Cen A. More broadly, this program highlights the complexity of analyzing a dynamic binary astrophysical scene and the challenges associated with confirming short-period (∼few years) planet candidates identified without prior orbital constraints in direct imaging searches. This Letter is second in a series of two papers: Paper I discusses the observation strategy and presents the astrophysical case (physical and orbital properties) for S1 as a planet candidate.