Confirmation of Iron Emission Lines and Non-detection of TiO on the Dayside of KELT-9b with MAROON-X

Abstract

We present dayside thermal emission observations of the hottest exoplanet KELT-9b using the new MAROON-X spectrograph. We detect atomic lines in emission at 10σ confidence using cross correlation with binary masks. The detection of emission lines confirms the presence of a thermal inversion in KELT-9b's atmosphere. We also search for TiO and other molecules, which have been invoked to explain the unusual HST /WFC3 spectrum of the planet. We do not detect any molecules, and instead use a retrieval approach to place an upper limit on the TiO volume mixing ratio of 10 −8.5 (at 99% confidence). This upper limit is inconsistent with the models used to match the WFC3 data, which require at least an order of magnitude more TiO, thus suggesting the need for an alternate explanation of the space-based data. Our retrieval results also strongly prefer an inverted temperature profile and atomic/ion abundances largely consistent with the expectations for a solar composition gas in thermochemcial equilibrium. The exception is the retrieved abundance of Fe + , which is about 1-2 orders of magnitude greater than predictions. These results highlight the growing power of highresolution spectrographs on large ground-based telescopes to characterize exoplanet atmospheres when used in combination with new retrieval techniques.

High-resolution Emission Spectroscopy of the Ultrahot Jupiter KELT-9b: Little Variation in Day- and Nightside Emission Line Contrasts

“…We robustly detected neutral iron in emission, confirming the previous detections of Fe I emission at highresolution by Pino et al (2020Pino et al ( , 2022 and Kasper et al (2021).…”

“…We found no evidence for emission by the molecular species TiO, in line with Kasper et al (2021) who derived an upper limit on the TiO VMR that is inconsistent with the VMR reported by Changeat & Edwards (2021).…”

“…However, their retrieved abundance of TiO suggests a super-solar metallicity. Kasper et al (2021) constrain the abundance of Fe I to be roughly consistent with so-lar abundances. However, their constraints on the Fe II abundance imply that KELT-9b's atmospheric metallic- The indications from our data of there being little variation in emission line contrast between the dayand nightsides suggests that it may be possible to detect nightside Fe I emission from KELT-9b.…”

“…The authors suggest that disequilibrium processes such as horizontal or vertical mixing may play a role in distributing these molecules to the dayside. Kasper et al (2021) observed the emission spectrum of KELT-9b with MAROON-X on Gemini-North and investigated its atmospheric composition with a highresolution spectral retrieval, similar to that of Brogi & Line (2019). Their retrieval considered H − , Fe I, Fe II, Ti I, Ti II, Ca II, Mg I, and TiO.…”