Nucleus of M31: Upper limits to the molecular and ionised gas content

We report observations performed with the NOrthern Extended Millimeter Array (NOEMA) and Atacama Large Millimeter/submillimeter Array (ALMA) of the nucleus of Andromeda (M31) that place strong constraints on the presence of gas in the cold or warm phase. M31 hosts the largest supermassive black hole (SMBH) closer than 1 Mpc to us. Its nucleus is silent, with some murmurs at the level of 4 × 10−9LEdd, and it is surrounded by a disc of old stars with a radius of 5 pc. The mass loss from these stars is expected to fill a molecular gas disc within the tidal truncation of 1 pc ( = 0.26 arcsec) of 104 M⊙, corresponding to a CO(1–0) signal of 2 mJy with a line width of 1000 km/s. We observed the nucleus with NOEMA in CO(2–1) and with ALMA in CO(3–2) with angular resolutions of 0.5″(1.9 pc) and 0.12″(0.46 pc), respectively. We exclude the presence of molecular gas with an upper limit of 3σ on the H2 mass of 195 M⊙ based on CO(3–2) ALMA observations. The CO(3–2) upper limit also constrains warm gas, which escapes detection in CO(1–0). The scenario of cold gas accumulation next to the nucleus of M31 that originates from mass loss of the old stellar population is not verified and excluded at a level of 150σ. The hot gas expelled by the stellar winds might instead never cool or fall onto the disc. Alternatively, the mass-loss rate of the stellar wind may have been overestimated by a factor 50, and/or the ionised gas has escaped from the nucleus. The SMBH in M31 clearly is in a low activity state, similar to what is observed for Sgr A* in the Milky Way (MW). Recently, a cool (104 K) ionised accretion disc has been detected around Sgr A* in the H30α recombination line with ALMA. If the sizes, masses, and fluxes were rescaled according to the mass of the black hole of M31 (35 times higher than in the MW) and its distance (97 times further away than in the MW), a similar disc might easily be detectable around the nucleus of M31. The expected signal would be eight times weaker that the signal detected in SgrA*. We searched for an ionised gas disc around the nucleus of M31 with NOEMA, and we place a 3σ upper limit on the H30α recombination line at a level twice lower than expected with a simple scaling of the SgrA*.