Promise and Peril: Stellar Contamination and Strict Limits on the Atmosphere Composition of TRAPPIST-1c from JWST NIRISS Transmission Spectra

Attempts to probe the atmospheres of rocky planets around M dwarfs present both promise and peril. While their favorable planet-to-star radius ratios enable searches for even thin secondary atmospheres, their high activity levels and high-energy outputs threaten atmosphere survival. Here, we present the 0.6--2.85\,$\mu$m transmission spectrum of the 1.1\,$\rm R_\oplus$, $\sim$340\,K rocky planet TRAPPIST-1\,c obtained over two JWST NIRISS/SOSS transit observations. Each of the two spectra displays 100--500\,ppm signatures of stellar contamination. Despite being separated by 367\,days, the retrieved spot and faculae properties are consistent between the two visits, resulting in nearly identical transmission spectra. Jointly retrieving for stellar contamination and a planetary atmosphere reveals that our spectrum can rule out hydrogen-dominated, $\lesssim$300$\times$ solar metallicity atmospheres with effective surface pressures down to 10\,mbar at the 3-$\sigma$ level. For high-mean molecular weight atmospheres, where O$_2$ or N$_2$ is the background gas, our spectrum disfavors partial pressures of more than $\sim$10\,mbar for H$_2$O, CO, NH$_3$ and CH$_4$ at the 2-$\sigma$ level. Similarly, under the assumption of a 100\% H$_2$O, NH$_3$, CO, or CH$_4$ atmosphere, our spectrum disfavors thick, $>$1\,bar atmospheres at the 2-$\sigma$ level. These non-detections of spectral features are in line with predictions that even heavier, CO$_2$-rich, atmospheres would be efficiently lost on TRAPPIST-1\,c given the cumulative high-energy irradiation experienced by the planet. Our results further stress the importance of robustly accounting for stellar contamination when analyzing JWST observations of exo-Earths around M dwarfs, as well as the need for high-fidelity stellar models to search for the potential signals of thin secondary atmospheres.