Molten Lithium-Brass/Zinc Chloride System as High-Performance and Low-Cost Battery

Abstract

•A high-performance and low-cost battery is reported for grid energy storage•Molten ZnCl2 cathode with high stability is realized by using brass in cathode•Using the low-cost ZnCl2 cathode reduces the cell cost to only $16 kWh−1•1-Wh-level SELL-brass/ZnCl2 full cell is successfully assembled and cycled Grid energy storage, which can solve the fluctuation and intermittence problems of wind and solar energies, is important for the penetration rate increase of the renewable energies in the grid. Batteries with high safety, low cost, and high energy density are urgently needed for grid energy storage. Molten Na-NiCl2 (ZEBRA) battery has being considered as one of the most promising candidates. However, using the expensive NiCl2 cathode results in a high battery cost, and replacing the NiCl2 cathode with low-cost metal chloride cathodes is difficult because of the poor stability of the low-cost cathodes. Here, by using brass (Cu-Zn alloy) instead of pure Zn, we significantly improved the stability of low-cost ZnCl2 cathode, based on which we successfully developed an SELL-brass/ZnCl2 battery. In addition to the advantages of high safety and high energy density, our SELL-brass/ZnCl2 battery also has a much lower cost; therefore, it has a high potential for grid energy storage application. Batteries with high safety, low cost, and reasonable energy density are essential for grid-scale energy storage and still remain elusive. Here, we report a solid electrolyte-based liquid lithium-brass/zinc chloride (SELL-brass/ZnCl2) battery using garnet-type lithium-ion solid electrolyte, lithium anode, and brass/ZnCl2 cathode. The chemistry of the cell reaction and the ability of being assembled in discharged state ensures a high safety. The use of low-cost ZnCl2 cathode can realize a low cell material cost of $16 kWh−1. The adoption of lithium anode guarantees a high theoretical energy density of 750 Wh kg−1 and 2,250 Wh L−1. Moreover, by using brass powder as a Zn source in the cathode, the Zn particle growth issue is successfully solved, and a good cycling stability of the battery can be obtained. As full cell performance and scalability are also verified, our SELL-brass/ZnCl2 battery shows a high potential for practical use in grid energy storage. Batteries with high safety, low cost, and reasonable energy density are essential for grid-scale energy storage and still remain elusive. Here, we report a solid electrolyte-based liquid lithium-brass/zinc chloride (SELL-brass/ZnCl2) battery using garnet-type lithium-ion solid electrolyte, lithium anode, and brass/ZnCl2 cathode. The chemistry of the cell reaction and the ability of being assembled in discharged state ensures a high safety. The use of low-cost ZnCl2 cathode can realize a low cell material cost of $16 kWh−1. The adoption of lithium anode guarantees a high theoretical energy density of 750 Wh kg−1 and 2,250 Wh L−1. Moreover, by using brass powder as a Zn source in the cathode, the Zn particle growth issue is successfully solved, and a good cycling stability of the battery can be obtained. As full cell performance and scalability are also verified, our SELL-brass/ZnCl2 battery shows a high potential for practical use in grid energy storage. The intrinsic fluctuation and intermittence of wind and solar energies have seriously limited their penetration rate in power systems, and grid-scale energy storage is urgently needed to stabilize output of the renewable energies before extensively integrating them into grids.1Chu S. Majumdar A. Opportunities and challenges for a sustainable energy future.Nature. 2012; 488: 294-303Crossref PubMed Scopus (5754) Google Scholar, 2Wang J. Conejo A.J. Wang C. Yan J. Smart grids, renewable energy integration, and climate change mitigation - future electric energy systems.Appl. Energy. 2012; 96: 1-3Crossref Scopus (59) Google Scholar, 3Suberu M.Y. Mustafa M.W. Bashir N. Energy storage systems for renewable energy power sector integration and mitigation of intermittency.Renew. Sust. Energy Rev. 2014; 35: 499-514Crossref Scopus (317) Google Scholar, 4Castillo A. Gayme D.F. Grid-scale energy storage applications in renewable energy integration: a survey.Energy Convers. Manage. 2014; 87: 885-894Crossref Scopus (377) Google Scholar Electrochemical energy storage is one of the important grid energy storage techniques, and batteries with high safety, low cost, and reasonable energy density are essential for grid-scale electrochemical energy storage, but they still remain elusive.5Gür T.M. Review of electrical energy storage technologies, materials and systems: challenges and prospects for large-scale grid storage.Energy Environ. Sci. 2018; 11: 2696-2767Crossref Google Scholar, 6Chu S. Cui Y. Liu N. The path towards sustainable energy.Nat. Mater. 2017; 16: 16-22Crossref Scopus (2073) Google Scholar, 7Dunn B. Kamath H. Tarascon J.M. Electrical energy for the grid: a choice of battery.Science. 2011; 334: 928-935Crossref PubMed Scopus (9183) Google Scholar, 8Luo X. Wang J. Dooner M. Clarke J. Overview of current development in electrical energy storage technologies and the application potential in power system operation.Appl. Energy. 2015; 137: 511-536Crossref Scopus (2096) Google Scholar The β″-Al2O3 Na-ion ceramic electrolyte-based molten Na-NiCl2 (ZEBRA) battery is considered as one of the most promising candidates.9Hueso K.B. Armand M. Rojo T. High temperature sodium batteries: status, challenges and future trends.Energy Environ. Sci. 2013; 6: 734-749Crossref Scopus (500) Google Scholar,10Lu X. Xia G. Lemmon J.P. Yang Z. Advanced materials for sodium-beta alumina batteries: status, challenges and perspectives.J. Power Sources. 2010; 195: 2431-2442Crossref Scopus (420) Google Scholar It has been extensively studied over the past decades and significant progress has been made.11Sudworth J.L. The sodium/nickel chloride (ZEBRA) battery.J. Power Sources. 2001; 100: 149-163Crossref Scopus (227) Google Scholar, 12Dustmann C.H. Advances in ZEBRA batteries.J. Power Sources. 2004; 127: 85-92Crossref Scopus (192) Google Scholar, 13Li G. Lu X. Kim J.Y. Meinhardt K.D. Chang H.J. Canfield N.L. Sprenkle V.L. Advanced intermediate temperature sodium-nickel chloride batteries with ultra-high energy density.Nat. Commun. 2016; 7: 10683Crossref PubMed Scopus (73) Google Scholar As assembled in fully discharged state, expensive and dangerous Na metal is not needed in the anode when assembling; however, a large amount of high-cost Ni powder is required in the cathode, which still hinders wide application of the ZEBRA battery.14Galloway R.C. Dustmann C.H. ZEBRA Battery - Material Cost, Availability and Recycling. MES-DEA GmbH, 2003Google Scholar, 15Lu X. Li G. Kim J.Y. Lemmon J.P. Sprenkle V.L. Yang Z. A novel low-cost sodium-zinc chloride battery.Energy Environ. Sci. 2013; 6: 1837-1843Crossref Scopus (36) Google Scholar, 16Li G. Lu X. Kim J.Y. Viswanathan V.V. Meinhardt K.D. Engelhard M.H. Sprenkle V.L. An advanced Na-FeCl2 ZEBRA battery for stationary energy storage application.Adv. Energy Mater. 2015; 5: 1500357-1500363Crossref Scopus (45) Google Scholar Efforts to replace Ni powder with low-cost metal (such as Fe, Al, Zn, and Cu) powders have long been plagued with difficulties.17Ratnakumar B.V. Attia A.I. Halpert G. Alternate cathodes for sodium-metal chloride batteries.J. Electrochem. Soc. 1991; 138: 883-884Crossref Scopus (18) Google Scholar, 18Ratnakumar B.V. Attia A.I. Halpert G. Sodium/metal chloride battery research at the Jet Propulsion Laboratory (JPL).J. Power Sources. 1991; 36: 385-394Crossref Scopus (19) Google Scholar, 19Moseley P.T. Bones R.J. Teagle D.A. Bellamy B.A. Hawes R.W.M. Stability of beta alumina electrolyte in sodium/FeCl2 (zebra) cells.J. Electrochem. Soc. 1989; 136: 1361-1368Crossref Scopus (42) Google Scholar Unlike Ni, the solubility of the low-cost metals' charge products (metal chlorides) in the molten catholyte (NaAlCl4) is high. This will benefit the utilization ratio of the metal powders during cycling, as there will be no insulating metal chloride precipitate coating on the metal particles to prevent the inner part of the particles from electrochemical reactions. However, the high solubility results in issues of metal particle growth through the Ostwald ripening mechanism and cationic ion exchange between the catholyte and β″-Al2O3 ceramic electrolyte during cycling, which cause rapid deterioration of battery performance and failure of the ceramic electrolyte. These issues have created the great challenge of replacing Ni with low-cost metals to develop a satisfactory ZEBRA-type battery for grid-scale energy storage applications. Here we report a solid electrolyte-based liquid lithium-brass/zinc chloride (SELL-brass/ZnCl2) battery, providing an effective strategy to overcome the issues. This SELL battery concept of molten Li metal anode with Li-ion solid electrolyte was first demonstrated by us in 2018 in systems using SELL-Sn-Pb and SELL-Bi-Pb,20Jin Y. Liu K. Lang J. Zhuo D. Huang Z. Wang C.A. Wu H. Cui Y. An intermediate temperature garnet-type solid electrolyte-based molten lithium battery for grid energy storage.Nat. Energy. 2018; 3: 732-738Crossref Scopus (121) Google Scholar later extended to SELL-S and SELL-Se.21Jin Y. Liu K. Lang J. Jiang X. Zheng Z. Su Q. Huang Z. Long Y. Wang C.A. Wu H. Cui Y. High-energy-density solid-electrolyte-based liquid Li-S and Li-Se batteries.Joule. 2020; 4: 262-274Abstract Full Text Full Text PDF Scopus (54) Google Scholar These reported systems confirmed the feasibility of the SELL battery concept, and, by using different cathode materials, the SELL batteries exhibit different advantages in terms of volumetric energy density, gravimetric energy density, rate capability, manufacturing cost, energy cost, and reliability. Among the possible cathode materials, as mentioned above, low-cost metal chloride cathodes are very attractive as they have reasonable energy densities, and combine the advantages of low energy cost of the S and Se cathodes, and high safety and good scalability of the alloy cathodes.9Hueso K.B. Armand M. Rojo T. High temperature sodium batteries: status, challenges and future trends.Energy Environ. Sci. 2013; 6: 734-749Crossref Scopus (500) Google Scholar, 10Lu X. Xia G. Lemmon J.P. Yang Z. Advanced materials for sodium-beta alumina batteries: status, challenges and perspectives.J. Power Sources. 2010; 195: 2431-2442Crossref Scopus (420) Google Scholar, 11Sudworth J.L. The sodium/nickel chloride (ZEBRA) battery.J. Power Sources. 2001; 100: 149-163Crossref Scopus (227) Google Scholar,15Lu X. Li G. Kim J.Y. Lemmon J.P. Sprenkle V.L. Yang Z. A novel low-cost sodium-zinc chloride battery.Energy Environ. Sci. 2013; 6: 1837-1843Crossref Scopus (36) Google Scholar,16Li G. Lu X. Kim J.Y. Viswanathan V.V. Meinhardt K.D. Engelhard M.H. Sprenkle V.L. An advanced Na-FeCl2 ZEBRA battery for stationary energy storage application.Adv. Energy Mater. 2015; 5: 1500357-1500363Crossref Scopus (45) Google Scholar,22Li H. Yin H. Wang K. Cheng S. Jiang K. Sadoway D.R. Liquid metal electrodes for energy storage batteries.Adv. Energy Mater. 2016; 6: 1600483-1600501Crossref Scopus (87) Google Scholar However, the ion exchange and metal particle growth issues should be addressed before they can be used in SELL batteries for practical applications. In this work, the key innovation of the SELL-brass/ZnCl2 battery to be demonstrated are: (1) by using brass powder as Zn source in cathode starting materials, the Zn particle growth is successfully suppressed via reversible copper-zinc alloying and de-alloying reactions during discharging and charging, respectively; and (2) by adopting Li anode and stable garnet-type Li6.4La3Zr1.4Ta0.6O12 (LLZTO) solid electrolyte, the ion exchange problem is significantly alleviated. Therefore, these strategies make a low-cost metal chloride cathode practical in a SELL battery system, and the SELL-brass/ZnCl2 battery could be a strong candidate for grid-scale energy storage. Figure 1 illustrates the structure of the SELL-brass/ZnCl2 battery, which is assembled in discharged state. A home-made garnet-type LLZTO solid electrolyte tube with high ionic conductivity, high relative density, and high stability is used, which allows fast Li-ion immigration between cathode and anode, while physically and electronically separating the two electrodes.20Jin Y. Liu K. Lang J. Zhuo D. Huang Z. Wang C.A. Wu H. Cui Y. An intermediate temperature garnet-type solid electrolyte-based molten lithium battery for grid energy storage.Nat. Energy. 2018; 3: 732-738Crossref Scopus (121) Google Scholar,21Jin Y. Liu K. Lang J. Jiang X. Zheng Z. Su Q. Huang Z. Long Y. Wang C.A. Wu H. Cui Y. High-energy-density solid-electrolyte-based liquid Li-S and Li-Se batteries.Joule. 2020; 4: 262-274Abstract Full Text Full Text PDF Scopus (54) Google Scholar,23Liu K. Ma J.T. Wang C.A. Excess lithium salt functions more than compensating for lithium loss when synthesizing Li6.5La3Ta0.5Zr1.5O12 in alumina crucible.J. Power Sources. 2014; 260: 109-144Crossref Scopus (71) Google Scholar Inside the LLZTO tube is the Li metal anode. As assembled in discharged state, no lithium metal is needed while assembling. To facilitate Li-ion transport, a small amount of LiI-CsI eutectic mixture with a melting point of 215°C is added inside the tube, serving as an anolyte. A roll of stainless steel mesh, as a part of the anode current collector, is also inserted. It can help form a layer of LiI-CsI melt on the surface of the inner wall of the LLZTO tube, thus it enlarges and stabilizes the active area for Li-ion immigration between the tube and the anode. Outside the LLZTO tube is the cathode, consisting of LiCl, brass (45 wt % Zn, 55 wt % Cu), and LiAlCl4 powders, serving as Li source, Zn source, and catholyte, respectively. The potential of Fe2+/Fe redox is higher than Zn2+/Zn redox; therefore, stainless steel should be stable within the electrochemical window of the ZnCl2 cathode. This allows us to use stainless steel as the cathode case and current collector, and also enables an anode-in-tube battery structure, which has a much higher cell capacity than a cathode-in-tube structure for a given ceramic electrolyte tube,24Oshima T. Kajita M. Okuno A. Development of sodium-sulfur batteries.Int. J. Appl. Ceram. Technol. 2004; 3: 269-276Google Scholar,25Wen Z. Cao J. Gu Z. Xu X. Zhang F. Lin Z. Research on sodium sulfur battery for energy storage.Solid State Ionics. 2008; 179: 1697-1701Crossref Scopus (150) Google Scholar thus realizing a low energy cost of the LLZTO tube (Table S1, Note S2 and Table S2). After assembling, the SELL-brass/ZnCl2 battery was cycled at an intermediate temperature of 250°C, which is about 50°C lower than the operating temperature of ZEBRA battery. In the first charge of the SELL-brass/ZnCl2 battery (Figure 2A), several tiny plateaus at about 2.15–2.25 V, and a major plateau at about 2.33 V can be seen. However, in the first discharge, there are two major plateaus in total: one at 2.12 V and another at 2.22 V. It seems that the charge and discharge plateaus in the first cycle do not correspond to each other. However, in the second charge, the curve differs from the first one significantly, and comprises two major plateaus at 2.18 V and 2.32 V, respectively, which matches well with the first discharge. To study the cell reactions, six samples were taken from cathodes at different stages of the cycling, which are shown in Figure 2A and labeled as Ⅰ to Ⅵ, respectively. The X-ray diffraction (XRD) patterns of the samples are shown in Figure 2B. At the initial stage (sample Ⅰ), the XRD peaks ascribe to the raw materials: LiCl, brass (comprising two phases: α and γ), and LiAlCl4 (Figure S1). As the first charge curve is not representative and the process cannot be divided clearly and meaningfully according to the curve, here we analyze the cell reactions starting from the end of the first charge (sample Ⅱ). Sample Ⅱ is at the fully charged stage, the peak intensity of LiCl significantly decreased compared with the raw materials, indicating a large part of LiCl in the cathode had been consumed. The brass became α phase with Cu, indicating the content of Zn element in the brass also significantly decreased, and the Zn element turned into ZnCl2. At the boundary of the two discharge plateaus (sample Ⅲ), the peaks of LiCl only slightly strengthened, and the ZnCl2 turned into Li2ZnCl4. The brass is mainly β phase, whose Zn content is higher than α phase. At the end of the first discharge (sample Ⅳ), the peaks of Li2ZnCl4 disappeared, the intensity of LiCl's peaks increased significantly, and the phases of the brass became α and γ again. Then at the boundary of the two charge plateaus (sample Ⅴ) and at the end of the second charge (sample Ⅵ), the patterns are very similar to that of sample Ⅲ and sample Ⅱ, respectively, which means a good reversibility of the battery reactions. In all of the six patterns, characteristic peaks of Zn were not observed, and the unmarked peaks belong to LiAlCl4, ZnCl2, or Li2ZnCl4 (Figure S1). Based on the XRD results of samples Ⅰ, Ⅱ, Ⅳ, and Ⅵ, we can conclude that the overall reaction of the battery is as follow:2Li+ZnCl2+CuxZny−1⇌2LiCl+CuxZny(Equation 1) The theoretical value of the open circuit voltage of a fully charged battery based on reaction (1) is estimated to be higher than 2.05 V (Note S1), and the experimental value (about 2.25 V) is consistent with the estimate. The XRD results of samples Ⅲ and Ⅴ show that Li2ZnCl4, with a melting point of 325°C,26Liu S. Liu Y. Zhang Q. The studies on the diagrams of the molten salt systems of ZnCl2-LiCl and SnCl2-ZnCl2-LiCl.Chem. J. Chinese U. 1984; 5: 765-768Google Scholar exists as an intermediate product during cycling. Accordingly, similar to the reaction process of a ZnCl2 cathode in a Na-ZnCl2 battery,27Lu X. Chang H.J. Bonnett J.F. Canfield N.L. Jung K. Sprenkle V.L. Li G. An intermediate-temperature high-performance Na-ZnCl2 battery.ACS Omega. 2018; 3: 15702-15708Crossref PubMed Scopus (8) Google Scholar the overall reaction can be divided into two reactions:Li+ZnCl2+CuxZny−1⇌1/2Li2ZnCl4+CuxZny−1/2(Equation 2) Li+1/2Li2ZnCl4+CuxZny−1/2⇌2LiCl+CuxZny(Equation 3) The first and second discharge plateaus can be assigned to reactions (2) and (3), respectively. Namely, at the first plateau, a large amount of free ZnCl2 exists in the cathode, which not only takes part in the electrochemical reaction but also combines with the resulting LiCl to form Li2ZnCl4 precipitate. Then, at the second plateau, where the free ZnCl2 is used up, the Li2ZnCl4 precipitate takes part in the reaction and free LiCl emerges and accumulates. This whole process matches well with the LiCl-ZnCl2 phase diagram (Figure S2), and the reverse process happens during charging; i.e., the first and second charge plateaus correspond to reactions (3) and (2), respectively. However, it should be noted that, according to reactions (2) and (3), the capacities of the two plateaus should be the same, but, in the tests, the capacity of the first discharge plateau is lower and the first charge plateau is higher. This could result from the influence of the molten LiAlCl4 catholyte, which can be considered as a mixture of AlCl3 and LiCl, and it can release a certain amount of LiCl to take part in the reactions. The extra LiCl will reduce the amount of free ZnCl2 and increase the amount of active LiCl at the beginning of discharge and charge, respectively, thereby decreasing the capacity of the first discharge plateau and increasing that of the first charge plateau. However, this influence does not affect the overall cell capacity and the cell reaction process does not change. To further investigate the behavior of the brass particles during cell reactions and the reason for the uniqueness of the first cycle, cathode samples Ⅰ to Ⅳ were dissolved in deionized water and filtered, therefore washing out the salts that covered the brass particles (Figure S3). Fe and Cu elements were not detected in the filtrates of the samples (Figure S4), which indicates a high stability of the stainless steel and the Cu component against electrochemical corrosion in the cathode. Figure 2C is the scanning electron microscopy (SEM) images of the brass particles filtered out from the cathode samples. As can be seen, the initially nonporous brass particles become porous after the first charging, and remain porous in the following cycling. The nonporous particles have lower specific area than the porous ones, which results in a higher internal resistance of the battery and a higher overpotential during cycling. Therefore, the dramatic morphology change of the brass particles in the first charge could be the reason that makes its plateaus abnormal, and further research on the morphology change is needed for better understanding of the process and for optimization of the morphology of the brass. Element content and distribution analysis of the brass particles by energy dispersive spectrometry (EDS) confirms that the Zn component is extracted out from the brass particles while charging and deposited back uniformly alloying with Cu while discharging (Figure S5), which matches well with the XRD results. The Cu component of the brass particles acts as stable frameworks to hold and release Zn through reversible Cu-Zn alloy and de-alloy reactions, respectively. The in situ-generated porous structure of the brass particles in the initial charge would benefit the cathode kinetics performance and consequently rate performance of the battery in the following cycles. The evolution of the LiCl and brass particles in the catholyte during cycling are summarized in Figure 2D. On the other hand, the size range of the brass particles in the samples is basically constant without obvious particle growth, which will also benefit the rate performance and cycling stability of the battery. As an explanation, instead of forming pure Zn particles, the produced Zn alloys with the stable Cu frameworks and is stored inside the brass particles; therefore, the Zn particle growth is avoided. Two control batteries replacing brass powder with graphite powder (named as SELL-graphite/ZnCl2 battery) and Zn powder (named as SELL-Zn/ZnCl2 battery), respectively, were assembled and tested. After only one cycle, Zn particles of the two batteries both grew much larger (Figure S6), and the capacities of the two batteries both decreased rapidly while cycling (Figure 3D). Since the graphite powder does not have any alloy reactions with Zn (Figure S7), and the Zn powder is not stable because it takes part in the electrochemical reactions, we can conclude that the reversible Cu-Zn alloy reaction and the inertness of the Cu component are essential to overcome the Zn particle growth problem in the system. To study battery performance of our SELL-brass/ZnCl2 battery, more SELL-brass/ZnCl2 batteries were assembled and tested, and the results are shown in Figure 3. As clearly shown in Figure 3A, after the first cycle, the overpotential significantly decreased and the cell capacity increased to a high level of 610 mAh g−1 (all of the specific capacities are calculated based on the weight of LiCl used in cathode), which means a lowered internal resistance of the battery and that the usage of LiCl increased from 73% to 97%. This activation process could be attributed to the formation of the porous structure of the brass particles after the first cycle. After being activated at 3.2 mA for three cycles, power rate and cycle life tests were conducted. As can be seen in Figures 3B and 3C, at a high current of 63.2 mA, the battery capacity is about 315 mAh g−1, which is 2.5 and 10.5 times that of the two control batteries, respectively (Figure S8). The average coulombic efficiency is higher than 99.9%, and the energy efficiency ranges from 85% to 94%, which are also at high levels. During the 100 cycles at 12.6 mA (0.2 C rate), the SELL-brass/ZnCl2 battery shows a stable cycling performance with average specific capacity, energy efficiency, and coulombic efficiency of 534 mAh g−1, 91.5%, and 99.99%, respectively (Figures 3D and S9). As a comparison, the two control batteries suffered from rapid capacity fade (Figure 3D). After 100 cycles, the cathodes of the SELL-brass/ZnCl2 battery and the control batteries were taken out and characterized. Results reveal continuous growth of Zn particles in both of the two control batteries (Figure S10), which could be the reason for their rapid capacity fade. As for the SELL-brass/ZnCl2 battery, the brass particles remained porous without obvious size change (Figures 3E and S10). The stability of the stainless steel and Cu component in the brass particles over a long operating time is also confirmed (Figure S11). The slow capacity fade after the 60th cycle could be attributed to the loss of LiAlCl4 catholyte resulting from volatilization of AlCl3 (Figure S11). We will develop a better sealing technique to address this problem in our future work. The LLZTO tube was characterized after being used in the SELL-brass/ZnCl2 battery for 100 cycles. Its XRD pattern was the same as that of a pristine LLZTO tube (Figure 3F), indicating a good crystal structure stability of the LLZTO tube. SEM images of the LLZTO tube before and after cycling (Figure S12) show that no obvious morphology change of the LLZTO tube can be found. Moreover, Zn element cannot be detected in the used LLZTO tube (Table S3), indicating a good stability of the LLZTO tube against Li-Zn ion exchange with the cathode. We suppose that, because the diameter of Li+ (68 pm) is much smaller than Na+ (98 pm), and no longer matches the size of Zn2+ (83 pm),19Moseley P.T. Bones R.J. Teagle D.A. Bellamy B.A. Hawes R.W.M. Stability of beta alumina electrolyte in sodium/FeCl2 (zebra) cells.J. Electrochem. Soc. 1989; 136: 1361-1368Crossref Scopus (42) Google Scholar the potential barrier for Zn2+ to penetrate LLZTO electrolyte and replace Li+ should be much higher than that in the case of β″-Al2O3 ceramic electrolyte. Therefore, the Li-Zn ion exchange in our SELL-brass/ZnCl2 battery can be effectively hindered. These results verify a high stability of the LLZTO tube in the battery system, which is of great importance for the stable performance of the battery. To investigate full cell performance and scalability of our SELL-brass/ZnCl2 battery, larger batteries with a discharge energy of 1 Wh were assembled and cycled. Notably, no Li metal was added in the anode while assembling. As shown in Figure 4A, at a high current of 31.6 mA, the energy efficiency and coulombic efficiency of the first cycle are 82.2% and 94.96%, respectively. These values are at relatively high level and indicate that side reaction in anode is minimal. Moreover, just like the small batteries, an activation process can be observed. In the following cycles, the discharge energy significantly increased from 0.79 Wh to 0.98 Wh. The energy efficiency and coulombic efficiency also increased to about 89.0% and 99.99%, respectively. Figure 4B is the charge-discharge curve of the 10th cycle. The plateaus and specific capacities match well with those of the small batteries shown in Figure 3C. As can be seen, at a scaled-up full cell state, the SELL-brass/ZnCl2 battery still shows good battery performance. We demonstrated a SELL-brass/ZnCl2 battery and solved the issues that have long plagued the development of high-performance and low-cost ZEBRA-type batteries. The strategy to overcome the metal particle growth issue via reversible alloying and de-alloying reaction could be more broadly applicable while developing other ZEBRA-type batteries. Our SELL-brass/ZnCl2 battery has a high theoretical energy density of 750 Wh kg−1 and 2,250 Wh L−1 (Table S4). By using an LLZTO tube with an inner diameter larger than 6.0 cm, single-cell-level practical energy densities of the SELL-brass/ZnCl2 battery can be higher than 250 Wh kg−1 and 750 Wh L−1 (Figure S13, Note S3, and Table S5), which are high enough for grid energy storage. More importantly, the cost of the cell materials (including electrodes, LLZTO tube, current collector, and battery case) of the SELL-brass/ZnCl2 battery is estimated to be $16 kWh−1, which is more than 50% lower than that of the ZEBRA battery (Table S6), and can be further reduced by optimizing the battery materials and battery structure. As the full cell performance and the scalability of the SELL-brass/ZnCl2 battery are verified, and considering its reasonable energy density, low cost, and high safety, we believe that the SELL-brass/ZnCl2 battery has a high potential for grid-scale energy storage application.