Ammonia is an indispensable component of modern fertilizers and a key nitrogen source in chemical synthesis, making it one of the highest-volume commodity chemicals.1–3 However, its synthesis via the Haber-Bosch process contributes significantly to global carbon emissions, motivating the development of alternative pathways for ammonia synthesis.4–6 Of many alternative methods, an electrochemical lithium-mediated process is one of the most promising. Since its initial implementations,7–9 novel cell designs10–12 and electrolytes13–15 have greatly improved the selectivity and productivity of lithium-mediated electrochemical ammonia synthesis (LiMEAS), but questions persist regarding the mechanism of nitrogen fixation and the role of lithium in this process (Figure 1a). In particular, many in the literature have alluded to the possible importance of lithium passivation species, commonly known as the solid electrolyte interphase (SEI), but their small length-scale and reactivity make direct observation challenging.11,13,15 In this work, we leveraged a multiscale approach to reveal connections between device-scale performance and nanoscale passivation behavior in LiMEAS. We used four model systems, varying the presence of proton donor (no proton donor or 0.1 M ethanol), and the feed gas (Ar or N2), which allowed us to examine the influence of the two reactive components in LiMEAS which are not usually present for lithium metal electrodeposition in the literature: proton donor and nitrogen. For each model system, we quantified major reaction products (Figure 1b, f). Ammonia and lithium nitride were measured by the salicylate colorimetric assay,16 hydrogen by online gas chromatography (GC), and residual lithium by galvanostatic stripping. We used scanning electron microscopy to observe microscale morphology, and to gain deeper insights into nanoscale surface morphology, we implemented cryo-electron microscopy (cryo-EM) (Figure 1c-d, g-h). Cryo-EM is a powerful technique that enables high-resolution observation of beam-sensitive metallic lithium and SEI materials by holding them at cryogenic temperatures, which preserve their native state.17 In addition to nanoscale morphological information, cryo-EM provided chemical information in the form of selected-area electron diffraction (SAED), energy-dispersive X-ray spectroscopy (EDS), and electron energy loss spectroscopy (EELS), which were paired with X-ray photoelectron spectroscopy (XPS) to develop an understanding of the chemistry of the interphase.Our results indicate that the proton donor governs reactivity toward nitrogen reduction in LiMEAS via its influence on the SEI. In the absence of proton donor, the lithium deposit forms a passivating, mosaic-structured SEI that preserves the deposited lithium, regardless of whether the feed gas is argon or nitrogen (Figure 1e). With the addition of ethanol, the surface changes significantly. The resultant SEI does not passivate the lithium, allowing it to react with nitrogen and other electrolyte components (Figure 1i). This permeable SEI appears to be mosaic in structure, with organic phases dominated by ethanol breakdown products, and its formation is key to the functionality of lithium-mediated ammonia synthesis. These findings can help guide the development of optimal SEIs for selectivity and stability of LiMEAS, and beyond that, offer insights into the characteristics of the lithium SEI at a reactive rather than passivated interface. Figure 1. (a) Schematic of possible lithium reactions. (b) and (f) Faradaic efficiencies of quantified reaction products, with no proton donor (b) and with ethanol present (f). (c-d) and (g-h) Cryo-TEM images of material deposited with no proton donor (c-d), and with ethanol present (g-h). (e) and (i) schematics of surface reactivity, with no proton donor (e) and with ethanol present (i). References J. W. Erisman, M. A. Sutton, J. Galloway, Z. Klimont, and W. Winiwarter, Nat. Geosci., 1, 636–639 (2008).G. R. Maxwell, Synthetic Nitrogen Products, Kluwer Academic Publishers, New York, (2004).FAO, World fertilizer trends and outlook to 2022, (2019).J. G. Chen et al., Science, 360 (2018).G. Soloveichik, Nat. Catal., 2, 377–380 (2019).Z. J. Schiffer and K. Manthiram, Joule, 1, 10–14 (2017).F. Fichter, P. Girard, and H. Erlenmeyer, Helv. Chim. Acta, 13, 1228–1236 (1930).A. Tsuneto, A. Kudo, and T. Sakata, Chem. Lett., 22, 851–854 (1993).A. Tsuneto, A. Kudo, and T. Sakata, J. Electroanal. Chem., 367, 183–188 (1994).N. Lazouski, M. Chung, K. Williams, M. L. Gala, and K. Manthiram, Nat. Catal., 3, 463–469 (2020).S. Z. Andersen et al., Energy Environ. Sci, 13, 4291–4300 (2020).K. Li et al., ACS Energy Lett., 30, 35 (2021).N. Lazouski et al., ACS Catal., 12, 5197–5208 (2022).B. H. R. Suryanto et al., Science, 372, 1187–1191 (2021).K. Li et al., Science, 374, 1593–1597 (2021).H. Verdouw, J. A. Van Echteld, and E. M. L. Dekkers, Water Res., 12, 399–402 (1977).Y. Li et al., Science, 358, 506–510 (2017). Figure 1