2024 Volume 92 Issue 11 Pages 117002
Sulfide all-solid-state batteries have been actively studied for practical use in vehicle applications. Modifications are often required at the interface between the sulfide solid electrolyte and oxide cathode active material. Scholars have reported that only the surface of the sulfide solid electrolyte is degraded by moisture exposure at a dew point equal to that of a dry room for common lithium-ion battery fabrication and that the surface-degraded material contains lithium carbonate and other lithium salts. Additionally, researchers have reported that lithium salts including lithium carbonate are effective for surface modification of cathode active materials. This paper reports how lithium carbonate is formed by the reaction of a carbon-free solid electrolyte with carbon-free water and that degraded surface of sulfide solid electrolyte by exposure to moisture acts as an effective modifying layer at the interface between the active material and solid electrolyte for all-solid-state batteries.
All-solid-state sulfide batteries (ASSBs) have been actively investigated for practical use in automotive applications. When sulfide solid electrolytes (SEs) are utilized in all-solid-state battery applications, a coating on the oxide cathode active materials is usually needed, as suggested by Ohta et al.1,2 Without the coating, the interface resistance is large and the battery does not perform satisfactorily. The most well-known coating layer is lithium niobate, another being lithium carbonate. Consequently, various studies have been conducted to explain the interface-modification effect mechanism.3–9
Some researchers, including our group, have reported that only the surfaces of sulfide SEs, including argyrodite-type sulfide SEs and Li3PS4-based SEs, are degraded by moisture exposure at a dew point equal to that of a typical drying room for lithium-ion battery production.10–20 We also reported that the surface-degraded material contains lithium carbonate and other lithium salts. Consequently, we hypothesized that such a surface-degraded layer could function as an interface-modification layer for use in sulfide ASSBs.21 There was also the question of why lithium carbonate is formed by the reaction of a carbon-free solid electrolyte with carbon-free water. In this regard, the humidified air originally contained about 500 ppm of carbon dioxide, and it was hypothesized that the carbon dioxide reacted.
In this study, we prepared air without CO2 and air containing CO2, adjusted the humidity of each, and exposed the sulfide solid electrolyte to the moisture at the dew point around −30 °C. The presence or absence of lithium carbonate formation on these samples was measured by X-ray photo electron spectroscopy (XPS) and the origin of lithium carbonate was investigated.
Next we exposed a solid sulfide electrolyte at the dew point around −20 °C, prepared samples whose surfaces were degraded, fabricated all-solid-state batteries using these samples, and investigated the effect of SE-surface degradation on the battery function. Although lithium nickel–manganese–cobalt oxide (NMC) has been widely studied as a cathode active material in recent years,22 scholars have noted that NMC may contain air-induced lithium residues such as hydroxide and carbonate residues on its surface,23,24 which can make detecting the effect of the modification-layer formation on the SE-side surface difficult. Therefore, lithium cobalt oxide (LCO)—which is less prone to forming degradation layers compared to NMC—was selected as the target sample. An argyrodite-type SE was selected as the sulfide SE.
As a result, the origin of the carbon on the degraded surface of the sulfide solid electrolyte was confirmed to be CO2 contained in the air. For the battery testing, a large overvoltage owing to the interfacial resistance at the sulfide SE/LCO interface was evident when pristine SE was used; we confirmed that the overvoltage was reduced considerably when degraded SE was used. The discharge capacity was also confirmed to be improved using the degraded SE.
An argyrodite-type SE, Li7−xPS6−xClx (x ∼ 1, D50 = 3.5 µm, Mitsui Mining and Smelting) was used as the sulfide-based SE. To investigate the effect of CO2 on moisture exposure, XPS measurements were conducted using ESCA-3400 (Shimadzu, MgKα radiation source) for the pelleted samples with the following two types of moisture exposure, where the dew point was adjusted to −30 °C and the exposure time was 42 h in both cases (Fig. 1a):
Schematic of the sample preparation for (a) XPS measurement and (b) battery test.
For the battery testing, the sulfide solid electrolyte was exposed to the moisture with CO2 and cells were fabricated following previous reports as schematically illustrated in Fig. 1b.9,25–27 The SE powder was exposed to humidity-controlled air with a dew point (d.p.) of −20 °C at a flow rate of 0.8 L min−1 for 1 h to prepare a moisture-exposed sample (hereinafter referred to as exposed SE, the non-exposed sample being referred to as pristine SE). Detailed material characterization was provided in our previous reports,16–18 revealing that the exposed SE surface contains decomposition products such as Li2CO3, as well as LiCl, Li2SO4, and Li3PO4. LiCoO2 (LCO, Nippon Chemical Industrial Co., Ltd.) was used without any surface treatment including a LiNbO3-coating as the positive electrode active material. The ASSB positive electrode half-cell was fabricated in an Ar glove box (d.p. less than −80 °C, O2 concentration < 1 ppm). The structure consisted of the following three layers, referencing previous reports:9,25,26
We examined the origin of the lithium carbonate on the sample exposed to the moisture air. The composition of the argyrodite-type SE did not contain C element. Additionally, the water that was considered to have reacted at the time of exposure did not contain C. Consequently, we compared the water exposure influence in two cases: one in which carbon dioxide was not actively included as an exposure condition and the other in which it was included. As described in experimental section, pellets of argyrodite-type SE were exposed using a gas that was prepared by dehumidifying ambient air (hereafter referred to “(gas-i) ambient gas with CO2”) and a gas that was prepared by exposing the glove box to the slight leakage of moisture in ambient air, with the glove box circulating compressed air without carbon dioxide (hereafter referred to “(gas-ii) gas cylinder without CO2”) (Fig. 1a). The surfaces of the obtained sample pellets were transported to the XPS chamber without exposure to a CO2-containing atmosphere, and the XPS C 1s spectra of the pellet surfaces were obtained. Figure 2 shows the XPS C 1s spectra of the samples exposed to the moisture with 500 ppm of CO2 (gas-i) and without CO2 (gas-ii). The spectra for the samples using gas-i and gas-ii have peaks at approximately 285 eV, which peaks are presumed to be derived from alkyl carbons, which are contaminants in the chambers. The spectrum of the gas-i sample has a peak at approximately 290 eV, whereas the spectrum of the gas-ii sample does not. This peak can be attributed to lithium carbonates,17 and the presence of CO2 in the environmental gas at the time of exposure contributed to this difference. The peak at 282 eV observed for gas-i sample is considered to be a ghost peak of the 290 eV peak. In this XPS measurement, a non-monochromatized MgKα source was used, which results in ghost peaks appearing at binding energy positions 8–10 eV lower. The conductivity of the pellets after exposure was measured, with the conductivity being 0.00014 mS/cm for the sample exposed to gas-i and 1.2 mS/cm for the sample exposed to gas-ii. This overwhelming difference in conductivity may also be related to the presence of lithium carbonate on the pellet surface and other salts, which decomposed product were detailly characterized in our previous reports.16–18
XPS spectra for the SE pellet samples exposed to (i) the moisture-controlled ambient gas with CO2 and (ii) the moisture-controlled gas cylinder without CO2.
Next, we investigated how such a denatured substance such as lithium carbonate affects battery operation by its presence at the interface between the active material and SE. Figure 3 shows the charge–discharge curves for the first three cycles of the cell using (a) a pristine SE and (b) an exposed SE. Figure 4 shows the (a) CC discharge capacity and CC/CV discharge capacity of each cell cycle obtained using the pristine SE and exposed SE and (b) Coulombic efficiency of CCCV capacity for these cells. In the case of the pristine SE cells, the overvoltage increases as the cycle progresses, and the charge–discharge capacity decreases accordingly. This charge–discharge behavior indicates that a side reaction occurs at the SE/LCO interface, and a resistance layer is generated in each cycle.26 In the case of the cell using the exposed SE, the capacity does not change appreciably during cycling. During the first charge, the charge capacity is slightly reduced; a slight irreversible capacity occurs, which is considered to be due to the oxidative decomposition of part of the SE when charged up to 4.3 V vs. Li.28 Both the CC and CC/CV discharge capacities are very stable, and almost no change occurs in the first three cycles, as shown in Fig. 3b. As shown in Fig. 4b, the Coulombic efficiency is higher for the cell using the exposed SE than that using the pristine SE. Furthermore, the overvoltage is considerably larger when the pristine SE is used than when the exposed SE is employed. This overvoltage trend can be attributed to the presence of lithium carbonate, lithium phosphate, and lithium sulfate on the surface of the exposed SE, but not on that of the pristine SE.17,18 It is considered that these Li compounds played a role in reducing the reaction resistance between SE and the active material, similar to the Nb coating layer. At the same time, the surface of each SE particle was modified, increasing the ion transfer resistance between SE particles. However, the reason for the observed improvement in battery performance upon exposure of SE is thought to be that the reduction in reaction resistance between SE and the active material was significantly greater than the increase in ion transfer resistance between SE particles. We previously reported on similar and different experiments, in which pristine and exposed SEs and Nb-coated NMC were used to construct an ASSB cathode. In those experiments, using an exposed SE instead of a pristine SE resulted in slightly larger overvoltage in the charge–discharge test.29 The trends of the SE exposure effect on the overvoltage are opposite in the two cases due to the presence of Nb-coating layer. This is because, in the presence of Nb coating, the reaction resistance between SE and the active material is sufficiently reduced by the Nb coating. Therefore, it is considered that the surface modification of SE only results in an increase in ion transfer resistance between SE and the active material, as well as between SE and SE. Several studies have demonstrated that modifying the surface or entire sulfide SE—rather than the surface of LCO—can lead to favorable battery performance, even with uncoated positive electrode active materials.30–37 The results of the present study demonstrate that the improved battery performance with the exposed SE is likely due to the same effect as reported in those studies about SE modification; the exposure of the SE was proven to be pseudo surface modification. As reported by Tsukasaki et al.,18 lithium carbonate is thought to be produced by the reaction of lithium hydroxide generated when the argyrodite-type SE reacts with water and carbon dioxide in air. Other researchers also proposed similar reactions in the presence of carbon dioxide.19,20 Some researchers reported that lithium halide hydrate prepared on the surface of the solid electrolyte suppresses the increase in resistance between the positive electrode and the solid electrolyte,36,37 and such lithium halide hydrate is quite possible to be formed on the surface of the solid electrolyte during exposure to humid air in the present study. These results provide the significant insight that surface degradation caused by exposure to humid air can induce a pseudo-surface modification effect on the sulfide SE.
Charge–discharge curves for the first three cycles for the cell using the positive electrode with (a) pristine SE and (b) exposed SE at 25 °C.
(a) (Black, red closed squares) CC and (black, red open circles) CCCV discharge capacities for the cell using a positive electrode with (black) pristine and (red) exposed SE. (b) Coulombic efficiency of the CCCV capacity for first three cycles of the cell using a positive electrode with (black) pristine and (red) exposed SE.
This study demonstrated that by exposing an SE to a moisture- and CO2-conditioned atmosphere, a pseudo surface-treatment is applied to the SE, with the formation of lithium carbonate and other lithium salts on the surface, which acts as a suitable modifier layer at the SE/LCO interface for the charge/discharge process. Additionally, this investigation demonstrated that lithium carbonate as a modified layer can be generated in the presence of CO2 in the environment during moisture exposure. Optimization of the modification, including exposing only the outermost surface and applying a heat recovery treatment, should be conducted in the future to develop more practical SE processes.
This study was conducted under the “Development of Fundamental Technologies for All-Solid-State Battery applied to Electric Vehicles (SOLiD-EV, JPNP18003)” project commissioned by the New Energy and Industrial Technology Development Organization (NEDO).
Hikaru Sano: Conceptualization (Lead), Data curation (Lead), Formal analysis (Lead), Investigation (Lead), Methodology (Lead), Project administration (Lead), Validation (Lead), Visualization (Lead), Writing – original draft (Lead), Writing – review & editing (Lead)
Yusuke Morino: Conceptualization (Equal), Data curation (Equal), Formal analysis (Equal), Investigation (Equal), Methodology (Equal), Validation (Equal), Visualization (Equal), Writing – review & editing (Equal)
Akihiro Shiota: Data curation (Equal), Formal analysis (Equal), Investigation (Equal), Validation (Equal)
Tsukasa Takahashi: Conceptualization (Equal), Data curation (Equal), Formal analysis (Equal), Investigation (Equal), Writing – review & editing (Equal)
Norihiko Miyashita: Data curation (Equal), Formal analysis (Equal), Investigation (Equal)
Koji Kawamoto: Conceptualization (Equal), Data curation (Equal), Funding acquisition (Equal), Project administration (Equal), Supervision (Lead), Visualization (Equal)
The authors declare no conflict of interest in the manuscript.
New Energy and Industrial Technology Development Organization: JPNP18003
H. Sano, Y. Morino, A. Shiota, T. Takahashi, N. Miyashita, and K. Kawamoto: ECSJ Active Members
