Context. Extreme adaptive optics systems, such as the Spectro-Polarimetric High-contrast Exoplanet REsearch facility (SPHERE), push forward the limits in high contrast and high resolution in direct imaging. The main objectives of these instruments are exoplanet detection and characterisation.
Aims: We aim to increase the contrast limits to detect new satellites orbiting known asteroids. We use cutting-edge data reduction techniques and data processing algorithms that are essential to best analyse the raw data provided by the instruments and increase their performances. Doing so, the unequalled performances of SPHERE also make it a unique tool to resolve and study asteroids in the solar system, expanding the domain of its main science targets.
Methods: We applied a newly developed data reduction pipeline for integral field spectrographs on archival SPHERE data of a resolved asteroid, (130) Elektra. It was coupled with a dedicated point spread function reconstruction algorithm to model the asteroid halo. Following the halo removal, the moon signal could be extracted more accurately. The moon positions were fitted at three epochs and were used to derive the orbital parameters via a genetic-based algorithm.
Results: We announce the discovery of S/2014 (130) 2, a third moon orbiting (130) Elektra, making it the first quadruple asteroid ever found. It is identified in three different epochs, 9, 30, and 31 Dec. 2014, at a respective angular separation of 258 mas (333 km), 229 mas (327 km), and 319 mas (457 km). We estimate that this moon has a period of 0.679 ± 0.001 day and a semi-major axis of 344 ± 5 km, with an eccentricity of 0.33 ± 0.05 and an inclination of 38° ±19° compared to the primary rotation axis. With a relative magnitude to the primary of 10.5 ± 0.5, its size is estimated to be 1.6 ± 0.4 km.
Conclusions: The orbital parameters of S/2014 (130) 2 are poorly constrained due to the unfavourable configurations of the available fragmentary data. Additional observations are needed to better estimate its orbit and to suggest a formation model. This new detection nonetheless shows that dedicated data reduction and processing algorithms modelling the physics of the instruments can push their contrast limits further.

Visualisations are available at https://www.aanda.org

Based on publicly available archival data provided by the ESO Science Archive Facility under programme ID 60.A-9362(A) (Yang et al.).