Nowadays, lithium metal anodes are often referred to as the ‘holy grail’ of next-generation battery technology. Compared to graphite anodes used in state-of-the-art lithium-ion batteries they promise several times higher energy densities [1]. However, Li metal shows a high reactivity with liquid electrolytes and uncontrolled dendrite growth which hinders a safe and efficient cyclability and therefore the commercial application in rechargeable batteries. Thus, stabilization of the solid-electrolyte interphase (SEI) between liquid electrolyte and the lithium surface is the key challenge to be solved before bringing this promising battery technology to market. In order to suppress unwanted effects such as dendrite formation, persistent loss of active material and high kinetic losses, this layer of electrolyte decomposition products needs to be electrically insulating and should prevent direct contact of metal with electrolyte while ensuring a good ionic conductivity and mechanical stability.Such a beneficial SEI design requires detailed understanding of the fundamental formation mechanism and the resulting chemical SEI composition and structure depending on electrolyte composition. In literature, a large range of experimental methods are applied to characterize the SEI under various conditions [2]. These approaches are mainly descriptive and struggle to identify and understand the underlying mechanisms. Theoretical approaches like density functional theory (DFT), ab-initio molecular dynamics (MD) or classical MD are well suited to determine reaction mechanisms and to model the very beginning of SEI formation in the ns-range. However, they are too computationally expensive to model technical relevant length- and time scales [3].Therefore, this work focuses on overcoming these restrictions by developing an efficient multiscale-model with input from ab-initio calculations. The acceleration originates from focusing on rare events such as reaction or diffusion processes in a 3-dimensional kinetic Monte Carlo instead of considering the vibrational motions of single atoms like in molecular dynamics. Thereby, every process is represented by a probability derived from ab-initio calculations to ensure physical correctness. An additional coupling with efficient ordinary differential equations as shown in [4] ensures electroneutrality in the simulation box.Implementation of the degradation reactions for a 1.2M LiPF6/EC electrolyte obtained from molecular simulations (see Figure 1, appendix), allows to unravel the initial SEI formation, the emerging structure and its composition within the first µs. The simulation suggests an almost instantaneous formation of SEI species on the pristine Li metal within the first nanoseconds after immersion. Due to this early passivation layer the formation process slows down continuously until electron tunneling is blocked by SEI species and a first passivation of the lithium surface is reached. Interestingly, the formed passivation layer mainly consists of inorganic species such as LiF close to the lithium metal surface and Li2CO3 throughout the whole layer but only of minor organic (CH2OCO2Li)2-molecules, suggesting that organic species are mainly formed at larger time scales which is the reason for the typical layering structure of the SEI.In summary this novel approach allows unprecedented insight of initial SEI-formation and thereby extends the accessible insight, time- and length-scales compared to molecular simulations.