Current architectures and fabrication approaches of solid-state batteries

Conventional LIBs use porous electrodes (both cathode and anode) and a porous separator with the pores filled with a liquid electrolyte.^{Reference Knoche, Zinth, Schulz, Schnell, Gilles and Reinhart16} On the other hand, solid-state batteries are usually made of dense layers of electrodes and electrolytes (Figure 1a). The use of dense layers is not only beneficial for high volumetric energy density, but also a prerequisite for intimate interfacial contact and fast Li⁺ conduction. The cathode of solid-state batteries is usually a composite layer of an active material, an ionically conductive solid electrolyte, and electronically conductive carbon. Conductive carbon is excluded sometimes if the electronic conductivity of the active material is sufficiently high.^{Reference Kamaya, Homma, Yamakawa, Hirayama, Kanno, Yonemura, Kamiyama, Kato, Hama, Kawamoto and Mitsui17–Reference Zhang, Leichtweiß, Culver, Koerver, Das, Weber, Zeier and Janek19} A high mechanical strength for the dense cathodes is required to sustain any volume change of the active materials since there is no free space to accommodate such changes. A dense layer of the solid electrolyte is also required to separate the cathode from the anode, which is also considered a potential solution for Li dendrite suppression and enabling high-capacity Li anodes.^{Reference Monroe and Newman20,Reference Neudecker, Dudney and Bates21} However, challenges remain in understanding the formation of Li dendrites along the grain boundaries and voids in certain solid electrolytes.^{Reference Ren, Shen, Lin and Nan12–Reference Han, Yue, Zhu and Wang14,Reference Porz, Swamy, Sheldon, Rettenwander, Frömling, Thaman, Berendts, Uecker, Carter and Chiang22} If conventional anode materials such as graphite and Li₄Ti₅O₁₂ are used in solid-state batteries instead of Li metal, a composite electrode could also be used as the anode.

Figure 1.

Schematics of the architecture and fabrication approaches for solid-state batteries. (a) Schematic representation of a typical bulk-type solid-state battery. (b) Cold-pressing process for sulfide solid electrolyte-based solid-state battery. (c) Sintering process for oxide solid electrolyte-based solid-state battery.

The fabrication approach for solid-state batteries largely depends on the mechanical properties of the solid electrolyte being used. Sulfide electrolytes are compliant, and therefore both the composite electrode and the solid electrolyte layer can be prepared by cold-pressing (Figure 1b),^{Reference Sakuda, Hayashi and Tatsumisago23} and a lithium disc can be attached to the electrolyte with minimal pressure. However, oxide-based solid electrolytes generally have better mechanical strength,^{Reference Yu, Schmidt, Garcia-Mendez, Herbert, Dudney, Wolfenstine, Sakamoto and Siegel24} and high-temperature sintering (Figure 1c) is usually required to ensure good contact within the solid electrolyte and at the interface between the solid electrolyte and the electrode.

Current approaches for preparing solid-state batteries cannot be easily scaled up to industrial production, and the performance of most solid-state batteries is not as good as those of liquid-electrolyte batteries. Figure 1a summarizes the main challenges for achieving a high-performance solid-state battery, including achieving intimate interfacial contact between the electrolyte and active material particles in the cathode composite, achieving intimate interfacial contact between the cathode composite layer and electrolyte layer, and fabricating a thin electrolyte layer. In addition, the bottleneck of scaling up for industrial fabrication is also discussed.

Intimate contact between active material and solid-electrolyte particles in cathode composite

An ionically and electronically conductive network in the cathode composite is crucial for high-performance solid-state batteries. Unlike conventional batteries where liquid electrolytes easily infiltrate into a porous cathode layer to form an intimate liquid−solid contact, solid-state batteries face challenges arising from insufficient solid−solid contact, leading to large interfacial resistance between the active material and solid-electrolyte particles.

Various approaches have been reported to increase the interparticle contact area in the cathode composite. A straightforward approach is to use submicrometer particles by mechanical milling of the active material, conductive carbon, and solid electrolyte using a planetary ball-mill apparatus.^{Reference Nagao, Hayashi and Tatsumisago25,Reference Nagao, Hayashi and Tatsumisago26} Different from hand-ground cathode composite (Figure 2a), better contact was obtained with submicrometer particles, forming an effective ion- and electron-transport pathway (Figure 2b).^{Reference Choi, Jin, Park and Lim27} Further enhancement of the contact area can be achieved by using solution-processible components. Some of the prevailing options are soluble sulfide solid electrolytes.^{Reference Liu, Fu, Payzant, Yu, Wu, Dudney, Kiggans, Hong, Rondinone and Liang28} A solution of the sulfides is either mixed with an active material to form solid electrolyte-coated active material after removing the solvent,^{Reference Lin, Liu, Dudney and Liang29–Reference Choi, Park, Kim, Oh, Kwak, Lee and Jung31} or infiltrated into a porous cathode fabricated for conventional liquid-electrolyte LIBs.^{Reference Kim, Oh, Park, Choi, Nam, Lee, Lee and Jung32} Both approaches allow active materials and solid electrolytes to form an intimate contact, and thus a more efficient ion-transport pathway as well as enhanced performance compared to dry-mixed electrodes.

Figure 2.

Schematics of approaches to improve contact between the active material and solid electrolyte particles in a cathode composite. (a) Cathode composite with large particles and limited contact area. (b) Cathode composite with finer particles and larger contact area. Reprinted with permission from Reference Reference Choi, Jin, Park and Lim27. © 2015 Elsevier. (c) Schematics of fabrication approach of mixed conducting cathode composite using solution-processible components. Reprinted with permission from Reference Reference Han, Yue, Fan, Gao, Luo, Ma, Suo and Wang33. © 2016 American Chemical Society. (d) Transmission electron microscope characterization of as-synthesized cathode composite. (e) Schematic of the cathode−electrolyte interface before and after co-sintering with an auxiliary wetting agent. (f) Cross-sectional scanning electron microscope image and elemental mapping images of cathode composite after sintering. Reprinted with permission from Reference Reference Han, Yue, Chen, Zhao, Fan, Ma, Gao, Wang, Guo and Wang18. © 2018 Elsevier. Note: LCBO, Li_2.3–xC_0.7+xB_0.3–xO₃; LLZO, Li₇La₃Zr₂O₁₂; PVP, poly(vinylpyrrolidone).

In addition to ionic conduction, an electronic pathway should also be simultaneously improved. Therefore, a solution-processible conductive carbon precursor was proposed, where soluble poly(vinylpyrrolidone) was coprecipitated with an active material and a solid electrolyte followed by thermal carbonization (Figure 2c).^{Reference Han, Yue, Fan, Gao, Luo, Ma, Suo and Wang33} The mixed ionic and electronic conducting cathode composite shows intimate triple-phase contact in the nanoscale (Figure 2d), resulting in a high utilization of the active materials. For solid-oxide electrolytes with a low solubility in solvents, auxiliary liquid or solid wetting agents may be introduced to facilitate interfacial contact. Liquid electrolytes are great examples of such wetting agents, though the introduction of liquids compromises the unique properties of solid-state batteries.^{Reference Huo, Zhao, Sun, Du, Li and Guo34} Li₃BO₃ (LBO) is a solid wetting agent that has a melting point of 700°C, enabling improved interfacial contact between Li₇La₃Zr₂O₁₂ (LLZO) particles at a relatively low sintering temperature (790°C).^{Reference Ohta, Seki, Yagi, Kihira, Tani and Asaoka35} The introduced LBO can also minimize element mutual diffusion between the active material and the solid electrolytes, which takes place at >500°C by physically separating the two during sintering.^{Reference Park, Yu, Jung, Li, Zhou, Gao, Son and Goodenough36} To further improve the contact and separation, the wetting agent Li_2.3C_0.7B_0.3O₃ reacts with Li₂CO₃ and forms a solid-solution interphase of Li_2.3–xC_0.7+xB_0.3–xO₃ (LCBO) after sintering at 700°C (Figure 2e). The active material and solid-electrolyte particles are tightly glued together by the presence of LCBO (Figure 2f).^{Reference Han, Yue, Chen, Zhao, Fan, Ma, Gao, Wang, Guo and Wang18} Overall, an intimate interparticle contact is nontrivial to the formation of an ionically/electronically conductive network in a cathode composite and continues to call for more effective solutions.

Intimate contact between the cathode layer and solid-electrolyte layer

In addition to the contact within the cathode composite, contact between the cathode layer and electrolyte layer can also become problematic. Due to the excellent deformability of sulfide solid electrolytes, intimate contact is achieved by simple cold-pressing. However, a rigid oxide solid electrolyte can leave microscopic gaps between the cathode and the solid-electrolyte layers and cause huge interfacial resistance (Figure 3a). The interfacial contact can be improved by the previously mentioned wetting agent-aided co-sintering,^{Reference Ohta, Seki, Yagi, Kihira, Tani and Asaoka35} or by advanced sintering techniques such as spark plasma sintering that could lower the sintering temperature (680°C) and shorten processing time (10 min).^{Reference Aboulaich, Bouchet, Delaizir, Seznec, Tortet, Morcrette, Rozier, Tarascon, Viallet and Dollé37} However, element mutual diffusion inevitably occurs at these temperatures (>500°C) and induces resistive interphase formation.

Figure 3.

Schematics of approaches to improve the cathode-solid electrolyte interfacial contact. (a) Schematic representation of interlayer contact improvement by a surface modified layer. (b) Schematic representation of a Nb-modified interface, suppressing elemental diffusion and facilitating the contact area. Reprinted with permission from Reference Reference Kato, Hamanaka, Yamamoto, Hirayama, Sagane, Motoyama and Iriyama38. © 2014 Elsevier. (c) Schematic representation of solid-state battery design with a conductive polymer at the cathode−electrolyte interface. Reprinted with permission from Reference Reference Zhou, Wang, Li, Xin, Manthiram and Goodenough39. © 2016 American Chemical Society. (d) Schematic representation of preformed monolithic bilayer structure after sintering. (e) Elemental mapping image (La in red; S in green) shows sulfur distribution and interconnectivity of bilayer framework. Reprinted with permission from Reference Reference Fu, Gong, Hitz, McOwen, Li, Xu, Wen, Zhang, Wang, Pastel, Dai, Liu, Xie, Yao, Wachsman and Hu43. © 2017 Royal Society of Chemistry.

Insertion of auxiliary interlayers has been proposed to improve the interlayer contact without degrading interface stability. A thin Nb layer was deposited on the surface of a LLZO pellet by radio-frequency magnetron sputtering and thermal annealing to form an Li⁺-conductive Li-Nb-O layer (Figure 3b). The interfacial resistance between LiCoO₂ and LLZO decreased from 2600 Ω cm² to 150 Ω cm² after the introduction of the Li-Nb-O interlayer, which was attributed to the elimination of voids/cavities and the suppression of element diffusion at the interface.^{Reference Kato, Hamanaka, Yamamoto, Hirayama, Sagane, Motoyama and Iriyama38} Alternatively, an additional soft polymer electrolyte layer inserted between the cathode and solid-electrolyte layer (Figure 3c) can wet the oxide solid-electrolyte surface, leading to a large interlayer contact area.^{Reference Zhou, Wang, Li, Xin, Manthiram and Goodenough39}

In addition to the interlayer approach, a monolithic porous cathode-dense electrolyte bilayer architecture represents a promising strategy to provide intimate contact and continuous ion transport if the active materials in the continuous porous channels can maintain good contact with the electrolyte wall during charge/discharge cycles.^{Reference Ren, Liu, Shen, Lin and Nan40–Reference Fu, Gong, Hitz, McOwen, Li, Xu, Wen, Zhang, Wang, Pastel, Dai, Liu, Xie, Yao, Wachsman and Hu43} The dense electrolyte layer was fabricated by tape casting an LLZO powder slurry while the porous electrolyte structure in cathodes was fabricated by adding a “pore former” (e.g., poly[methyl methacrylate]) in the slurry using the same method.

Finally, the monolithic bilayer structure was formed by laminating and co-sintering (Figure 3d).^{Reference Fu, Gong, Hitz, McOwen, Li, Xu, Wen, Zhang, Wang, Pastel, Dai, Liu, Xie, Yao, Wachsman and Hu43} The continuous connected pore channels in the porous layer allow subsequent infiltration of carbon nanotube ink and melted sulfur as the active materials. The direct connection between the cathode and solid electrolyte layers was demonstrated through elemental mapping (Figure 3e [La, red; S, green]). The previously discussed approaches can also be extended to improve the interfacial contact between anode and solid electrolyte.^{Reference Han, Gong, Fu, He, Hitz, Dai, Pearse, Liu, Wang, Rubloff, Mo, Thangadurai, Wachsman and Hu44,Reference Yang, Zhang, Liu, Xu, Hamann, McOwen, Dai, Luo, Gong, Wachsman and Hu45}

Thin solid electrolyte fabrication

To achieve higher specific gravimetric energy (Wh kg⁻¹) and volumetric energy density (Wh L⁻¹) at the cell level than those of liquid-electrolyte LIBs, the solid electrolyte layer must be thinner than a critical value, or “break-even thickness.”^{Reference Wu, Wang, Evans IV, Deng, Yang and Xiao46,Reference McCloskey47} For example, for a cathode with 19.5 mg cm⁻² loading, the specific gravimetric energy break-even thickness for oxide (LLZO) and sulfide solid electrolyte layers (Li₁₀GeP₂S₁₂ [LGPS]) is 45 µm and 122 µm, respectively.^{Reference Wu, Wang, Evans IV, Deng, Yang and Xiao46} The sulfide-based solid electrolyte with a smaller mass density (LGPS, ρ ∼ 2 g cm⁻³; LLZO, ρ ∼ 5 g cm⁻³) shows a larger break-even thickness. Solid electrolytes with less than 120-µm thickness are necessary, which is still hard to reach using traditional fabrication. Thus, novel thin-electrolyte fabrication methods are essential.

Advanced techniques are emerging for fabricating thin oxide solid electrolytes with extremely conformal thin-coating quality such as pulsed laser deposition,^{Reference Kim, Hirayama, Taminato and Kanno48} sol-gel,^{Reference Chen, Huang, Huang, Shen, Lin and Nan49} aerosol deposition,^{Reference Ahn, Choi, Ryu, Hahn, Kim, Yoon, Choi and Park50} radio-frequency magnetron sputtering,^{Reference Lobe, Dellen, Finsterbusch, Gehrke, Sebold, Tsai, Uhlenbruck and Guillon51} and atomic layer deposition,^{Reference Kazyak, Chen, Wood, Davis, Thompson, Bielinski, Sanchez, Wang, Wang, Sakamoto and Dasgupta52} though some of these approaches can be high cost and difficult to scale up. A scalable fabrication of a thin LLZO film was demonstrated with casting-sintering flame-made LLZO nanopowders.^{Reference Yi, Wang, Kieffer and Laine53} While typical LLZO pellets prepared from micropowders require sintering at >1000°C, the nanopowder-derived film was sintered at 700−800°C and was considerably dense and thin (<30 µm). Stacking such brittle thin films without damage is a challenge.

The break-even thickness for sulfide-solid electrolytes is larger than that for oxide-solid electrolytes, and thus technically easier to achieve.^{Reference Wu, Wang, Evans IV, Deng, Yang and Xiao46} Therefore, sulfide-solid electrolytes could be more promising for high cell-level specific energy and energy density.^{Reference Inada, Kobayashi, Sonoyama, Yamada, Kondo, Nagao and Kanno54–Reference Oh, Kim, Jung, Han, Choi and Jung59} Thin sulfide-solid electrolytes can be fabricated with the help of a compliant polymer scaffold.^{Reference Nam, Cho, Oh, Lim, Kim, Song, Lee, Lee and Jung55} Figure 4a shows a self-standing and bendable thin sulfide solid electrolyte layer (70 µm) fabricated by coating an Li₃PS₄ slurry on an Ni foil, followed by drying and cold-pressing onto a polymer (poly[para-phenylene terephthalamide]) nonwoven scaffold, which provides flexibility and toughness. The solid-state battery assembled using this thin solid electrolyte delivered a threefold increase in cell-level specific energy compared with that of a conventional pelletized solid-state battery.^{Reference Nam, Cho, Oh, Lim, Kim, Song, Lee, Lee and Jung55} Introducing polymeric binders also provides the mechanical flexibility and adhesion between solid electrolyte particles necessary for achieving a thin layer.^{Reference Sakuda, Kuratani, Yamamoto, Takahashi, Takeuchi and Kobayashi56,Reference Lee, Kim, Park, Park, Coskun, Jung, Cho and Choi58} A thin solid electrolyte sheet (∼70 µm) was prepared by coating a slurry consisting of glassy Li₃PS₄ solid electrolyte and styrene-butadiene-styrene copolymer binder on a copper foil.^{Reference Sakuda, Kuratani, Yamamoto, Takahashi, Takeuchi and Kobayashi56} Since polymer binders are ionic insulating, a thermal decomposable poly(propylene carbonate) binder was designed for fabricating thin glassy Li₃PS₄ sheets (∼60 µm).^{Reference Yamamoto, Terauchi, Sakuda and Takahashi57} The binder could be fully removed by heat treatment at 225°C after stacking and compressing steps (Figure 4b). The solid-state battery thus formed exhibited cell-level specific energy of 115 Wh kg⁻¹, a 25% increase from that of the binder-containing solid-state battery (92 Wh kg⁻¹).^{Reference Oh, Kim, Jung, Han, Choi and Jung59}

Figure 4.

Fabrication approaches of solid-state batteries with thin solid electrolytes. (a) Schematic representation of fabrication of thin solid electrolyte with polymer scaffold. Reprinted with permission from Reference Reference Nam, Cho, Oh, Lim, Kim, Song, Lee, Lee and Jung55. © 2015 American Chemical Society. (b) Schematic representation of fabrication of binder-free styrene-butadiene-styrene. Reprinted with permission from Reference Reference Yamamoto, Terauchi, Sakuda and Takahashi57. © 2018 Nature Publishing Group.

Challenges for large-scale fabrication

Bringing solid-state batteries to the market requires scaling up from lab-scale fabrication to industrial production. In this section, we provide an overview of the main challenges for the large-scale fabrication of solid-state batteries.

Summary and outlook

We have reviewed architectures of solid-state batteries and highlighted fabrication approaches for achieving intimate interparticle and interlayer contact. The contact evolution upon cycling, however, is not well understood, and alternative materials, cathode configurations, and fabrication approaches may be needed to maintain excellent interfacial contact during battery cycling. The thin solid electrolyte film is critical for high cell-level specific energy and energy density. Thin solid electrolyte films are prepared from soluble solid electrolytes with the aid of polymer scaffolds or binders. However, not all solid electrolytes are soluble, and ionic insulating polymer scaffolds or binders can compromise ion conduction. Soluble solid electrolytes with high ionic conductivity and polymer binders removable at low temperatures are therefore highly desirable. For integration of thin solid-electrolyte films into solid-state batteries, a cathode-supported solid-state battery, similar to a cathode-supported solid oxide fuel cell,^{Reference Menzler, Tietz, Uhlenbruck, Buchkremer and Stöver73} may be developed. Future research for solid-state battery fabrication will be directed toward large-scale approaches; roll-to-roll fabrication and three-dimensional printing are worth considering in this regard.