Volatile cycles are crucial feedback processes balancing Earth´s chemical and physical systems. The outgassing of volatiles such as H-C-O (hydrogen, carbon, and oxygen) is an essential source for the formation of the atmosphere and hydro- sphere. These reservoirs significantly shaped the Earth´s surface and transformed our planet into a habitable environment. Especially on early Earth, hot mantle temperatures led to vigorous convection, resulting in enhanced volcanism and volatile outgassing. However, hydrous conditions in the mantle are proposed to be required to develop the continental crust, another major player in shaping the Earth´s surface and determining its evolution. This renders the recycling of volatiles a critical sink, rehydrating the mantle. Volatile cycles stabilize the global climate (by maintaining clement conditions) and influence physical and chemical mantle properties (e.g., oxygen fugacity, solidus temperature, viscosity, and den- sity). All these processes and feedback mechanisms render investigating volatile cycles essential for geology and natural sciences in general. In this thesis, I focus on three research questions regarding H-C-O cycling from an observational, theoretical, and modeling point of view: • How did the mantle redox state affect volatile outgassing on early Earth and vice versa? • What are the relative amounts of H2O and CO2 that can be released from an intrusive system on (early) Earth? • Which parameters influence the deep recycling of water, the melt generation, and the water-melt ascent? The Earth´s redox state (considered crucial for the emergence of life) is significantly influenced by volatile cycles and conversely. To answer the first research question of this thesis, I review the accretion and partitioning of volatiles on early Earth. Additionally, I investigate the H-C-O contribution to the early atmosphere. Here, I consider intrusive and extrusive volatile release, including the redox state of the melt, by combining my newly developed RICH (Release of Intrusive CO2 and H2O) code with the volatile speciation code of Caroline Brachmann. The results reveal that the proportions of outgassed volatiles can vary considerably for intrusive versus extrusive systems and for distinct oxygen fugacities. Finally, I discuss the mechanisms that have led to a delayed accumulation of free oxygen in Earth´s atmosphere (e.g., subaerial outgassing of oxidized or burial of reduced species). The primary source of Earth´s atmosphere is volatile outgassing (intrusive and extrusive). As previous publications have already investigated extrusive out- gassing, the second study of this thesis focuses in detail on intrusive volatile release and its potential contribution to the (early) atmosphere. For this purpose, I developed the RICH code that calculates the solubility, accumulation (due to fractional crystallization), and release of H2O and CO2 from an intrusive melt. Furthermore, this study considers the buoyancy of the melt and the uptake of OH via the formation of hydrous phases. I found that about 0 - 85 % of H2O and 100 % of CO2 can be released from intrusive systems depending on the initial volatile content, the buoyancy of the melt, and the formation of hydrous minerals. These results indicate that the total outgassing on Earth is likely underestimated if intrusive volatile release (including fractional crystallization and repartitioning of volatiles in the system) is neglected. Therefore, the process of fractional crystallization and the resulting release of volatiles from intrusions are crucial in estimating outgassing fluxes on rocky planets, including Earth. Closing the volatile cycle, recycling is the essential volatile sink process, returning volatiles to the Earth´s interior. To investigate the depth to which water can be recycled and water outgassing during subduction, I enhanced a 2D numerical convection code, simulating a subduction process. The innovative features of this third study are the coupling of the dehydration calculation with that of the melt generation and the introduction of a saturation efficiency factor. This factor controls the proportion of excess water reacting with the surrounding rock. One result of this study is that the saturation efficiency factor significantly influences the extent of the (water-saturated) melt ascent. If the water that reacts with the surrounding rock exceeds 40 %, the water-saturated melt does not reach the surface (within the maximum simulation time of 50 Myr). Furthermore, the ascent of the water-saturated melt is facilitated by comparatively high initial water contents, high dip angles, and low plate velocities. In contrast, the deep recycling of water is favored by low dip angles, high plate velocities, and comparatively low initial upper mantle temperatures. In summary, this thesis contributes to an enhanced understanding of volatile cycles by investigating the sources (outgassing) and sinks (recycling). Specifically, it reveals the importance of intrusive volatile release and the redox state for total outgassing on early Earth. Regarding volatile recycling, the thesis highlights the significance of the water-melt ascent as an essential process forming arc volcanism.
