The formation of the planetary cores of gas giants via the accretion of planetesimals takes very long and is not compatible with the lifetime of protoplanetary discs (Levison et al. 2010). This time-scale problem can be solved through the accretion of pebbles onto a planetary seed. Contrary to planetesimals, pebbles feel the headwind from the gas which robs them of angular momentum allowing an efficient growth from the entire Hill sphere, which reduces the growth time-scale by several orders of magnitude (Lambrechts & Johansen, 2012; 2014). However, pebble accretion self-terminates when the planets start to open a partial gap in the disc, which accelerates the gas outside of the planets orbit and thus stops the flow of pebbles onto the planetary core (Lambrechts et al. 2014). Typically this mass is of the order of 10-20 Earth masses, depending on the local disc properties. The planet can then start to accrete a gaseous envelope without a pollution of pebbles. During its growth, the planet migrates through the disc, which evolves in time (Bitsch et al. 2015a,b).

Different volatile species like CO2 or H2O have different condensation temperatures and are thus present in either solid or gaseous form at different locations in the disc. A pebble accreting planet can thus only accrete volatiles that are in solid form, while a gas accreting planet will only accrete volatiles which are in gaseous form. Therefore the final chemical composition of the planetary atmosphere of a giant planet is strongly influenced by the formation location of the initial planetary seed and its subsequent migration path through the disc. Additionally, the envelope can be enriched through the erosion of the planetary core.

I will discuss the implications of the formation of planets via pebble accretion and their subsequent migration through the disc on the composition of gas giants. In particular I will focus on the carbon to oxygen abundance of hot Jupiters as well as on Jupiter in our own solar system.​