ARTICLE DOI: 10.1038/s41467-018-06077-5 OPEN Solubility-mediated sustained release enabling nitrate additive in carbonate electrolytes for stable lithium metal anode Yayuan Liu 1, Dingchang Lin1, Yuzhang Li1, Guangxu Chen1, Allen Pei 1, Oliver Nix1, Yanbin Li1 & Yi Cui1,2 The physiochemical properties of the solid-electrolyte interphase, primarily governed by electrolyte composition, have a profound impact on the electrochemical cycling of metallic 1234567890():,; lithium. Herein, we discover that the effect of nitrate anions on regulating lithium deposition previously known in ether-based electrolytes can be extended to carbonate-based systems, which dramatically alters the nuclei from dendritic to spherical, albeit extremely limited solubility. This is attributed to the preferential reduction of nitrate during solid-electrolyte interphase formation, and the mechanisms behind which are investigated based on the structure, ion-transport properties, and charge transfer kinetics of the modified interfacial environment. To overcome the solubility barrier, a solubility-mediated sustained-release methodology is introduced, in which nitrate nanoparticles are encapsulated in porous poly- mer gel and can be steadily dissolved during battery operation to maintain a high con- centration at the electroplating front. As such, effective dendrite suppression and remarkably enhanced cycling stability are achieved in corrosive carbonate electrolytes. 1 Department of Materials Science and Engineering, Stanford University, Stanford, CA 94305, USA. 2 Stanford Institute for Materials and Energy Sciences, SLAC National Accelerator Laboratory, 2575 Sand Hill Road, Menlo Park, CA 94025, USA. Correspondence and requests for materials should be addressed to Y.C. (email: [email protected]) NATURE COMMUNICATIONS | (2018) 9:3656 | DOI: 10.1038/s41467-018-06077-5 | www.nature.com/naturecommunications 1 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-06077-5 he morphology of electrodeposited metals is dictated to a extremely low solubility. However, we observe that the preferential – Tgreat extent by the physiochemical characteristics at the reduction of NO3 during SEI formation in carbonates, albeit at metal/electrolyte interface, including ion transport, inter- very low concentration, could substantially alter the interfacial facial energy, and mechanics, etc.1. This especially stands for the chemistry, resulting in spherical Li nuclei (instead of dendritic) and electrochemical plating of metallic lithium (Li), the ultimate greatly improved CE exceeding the values in ethers. To further anode for next-generation batteries with highest specific capacity overcome the solubility limitation, a solubility-mediated sustained- –1 (3860 mAh g ) and lowest electrode potential of all possible release concept is proposed, where LiNO3 nanoparticles are dis- alternatives2,3. In particular, the electrodeposition of Li is com- persed in porous polymer backbone on anode surface and can be plicated by the instantaneous formation of a resistive interfacial steadily dissolved during battery operation when soluble LiNO is – 3 passivation (i.e. the solid-electrolyte interphase, SEI), originated consumed. With the maintained concentration of NO3 on Li from the parasitic reduction of electrolyte components by the surface frontier, the anode CE can exceed 98% for over 200 cycles, 4 highly reactive Li . The chemical heterogeneity and mechanical and the cycle life of full-cells paired with LiNi1/3Mn1/3Co1/3O2 instability of the SEI layer is generally believed to induce non- (NMC) is more than quadrupled, which is appreciable in corrosive uniform ion flux, resulting in the formation of Li dendrites that carbonate electrolytes. Note that during the peer-review process of could lead to internal short circuit and compromise battery safety, our manuscript, we noticed another work utilizing a slightly similar while the repeated breakdown and repair of SEI brings about engineering method to overcome the solubility barrier of LiNO3 in continuous loss of active materials, giving rise to limited battery carbonate electrolyte, which also resulted in improved Li anode cycle life5. CE29. Nevertheless, we are confident that our comprehensive work Extensive research has been devoted to regulate the surface brings a significant amount of new knowledge complementary to – reactivity of Li metal. Among all the tactics explored to date the other study. Particularly, the mechanism of NO3 in modifying (protective coatings6,7, nanostructured electrodes8,9, high- the SEI properties is studied deliberately on the basis of cryo- modulus separators10,11, etc.), tailoring the electrolyte composi- electron microscopy (cryo-EM)30,31, ultramicroelectrode, X-ray tion is among the most essential and prominent paradigms, for it photoelectron spectroscopy (XPS), and electrochemical impedance can directly impact the physiochemistries of the SEI layer, spectroscopy (EIS). The characterizations unravel a distinct struc- modifying the interfacial environment to alter Li deposition tural change of the SEI from amorphous to bilayered configuration, behavior12–16. a strong presence of nitrogen-containing species inside the SEI, and It is known that the choice of electrolyte solvents can already a greatly increased exchange current density of the Li+/Li couple – offer pronounced effect on Li deposition uniformity. In general, with the addition of NO3 , and the effects of which are discussed in ether-based electrolytes demonstrate relatively controlled detail. This work not only enables uniform Li deposition with high deposition and thus high Coulombic efficiency (CE), ascribed to reversibility in typical carbonate-based electrolytes without com- the formation of oligomeric SEI with superior flexibility17. The promising the stability, ionic conductivity, viscosity, and cost of the application of ether electrolytes, however, is severely hampered by electrolytes, but also provides fruitful insights on the exact role of its high flammability and low oxidation potential (anodic nitrate additive at a fundamental level. decomposition at <4 V vs Li+/Li)18, mismatched with the safety benchmark and the emerging high-voltage cathodes19. On the other hand, carbonate-based electrolytes exhibit lower flamm- Results ability and higher anodic potential, and, therefore, are used Effects of nitrate on Li morphology in carbonate electrolyte.As exclusively in almost all the commercial Li-ion batteries. Never- frequently mentioned in the literature, the cathodic decomposi- – + 32 theless, they are highly corrosive to metallic Li, rendering tion of NO3 can start as early as ~1.7 V vs Li /Li , which is aggravated dendrite formation and poor cycling efficiency. much higher than other carbonate-based electrolyte components, Fluorinated molecules, such as fluoroethylene carbonate (FEC), making it a promising candidate as “electrolyte-agnostic” additive 33 – are frequently adopted as film-forming additives in carbonates, (Fig. 1a) . Therefore, if the preferential reduction of NO3 can which is believed to promote the formation of LiF on Li surface as yield an SEI conducive to controlled Li deposition, the presence of 20–22 – a favorable SEI passivation component . Even with these NO3 in carbonate electrolytes is highly likely to afford improved additives, the Li anode CE remains deficient for practical appli- Li metal performance. To determine the exact reduction potential – cations, and the adverse effects of FEC decomposition at elevated of NO3 , cyclic voltammogram (CV) of 1.0 M LiNO3 was temperatures bring concerns when being used in full-cells23. High obtained on a stainless-steel electrode at a scanning rate of 0.1 salt concentration has recently arisen as another exemplary mV s–1. Dimethoxyethane (DME) was selected as the solvent for method to stabilize Li metal in carbonate electrolytes, the effect of the CV scan, due to its stability against reduction (–1.68 V vs Li + 34 which is originated from the unique solvation structure that /Li) and high LiNO3 solubility . As can be observed in Fig. 1b, reduces the solvent reactivity24–26. Nevertheless, its economic there is a deep cathodic current peak beginning at ~1.69 V vs Li/ effectiveness, rheological properties, and ion transport need to be Li+ in the first CV cycle, which disappeared upon the second evaluated for practical applications. Thus, it is apparent that the scan. And no similar reduction peak can be observed in the first development of new electrolyte reformulation strategies, which CV cycle when LiNO3 was replaced by lithium bis(tri- could circumvent the potential drawbacks of the existing ones, fluoromethanesulfonyl)imide (LiTFSI). Thus, the irreversible remains highly desirable. cathodic current starting at ~1.69 V vs Li/Li+ can be assigned Theoretically, if an additive is much more liable towards reduc- with confidence to the reduction of NO –. The easy reduction of – 3 tion than other electrolyte components to form an interfacial NO3 participates as the frontier reaction during SEI formation environment favorable for Li deposition, good electrochemical and plays a decisive role in the physiochemical properties of the performance can be achieved regardless of the electrolyte systems SEI. For this reason, the addition of LiNO3 in ether electrolytes (“electrolyte-agnostic” additive). Herein, we report the discovery of can drastically change the Li deposition morphology from irre- – the pronounced effect of nitrate anions (NO3 )onthemorphology gular to densely packed spheres, together