Characterization of All Solid State Batteries with LiPON Thin Films Obtained with Different Substrates and RF Sputtering Times
2018 年 59 巻 7 号 p. 1156-1160
詳細
2018 年 59 巻 7 号 p. 1156-1160
All-solid-state lithium batteries consisting of LiPON thin films were prepared by radio-frequency (RF) magnetron sputtering, and the influences of substrates and RF sputtering times have been investigated. For the preparation with different substrates, the Cu, Mo, Al and FTO (SnO2: F) were chosen to grow LiPON films, respectively. The results indicate that the small particles of FTO-sub film can aggregate together and form the particles clusters, which contribute to a fairly rough surface and high discharge capacity of the lithium cell during the cycle test. For the preparation with different sputtering times, the 0.5, 1.0, 1.5 and 2.0 hours are taken to evaluate the characteristics and electrochemical properties of the films, respectively. The results show that long sputtering time (1.5 and 2.0 h) could prevent the formation of polycrystalline phases, which may increase the capacity of all-solid-state Li+ battery.
Recently, the rechargeable thin-film Li+ battery, has drawn much interest due to the growing demand for advanced energy storage applications such as mobile electronics and electric vehicles.1–7) Effective rechargeable films Li+ ion batteries typically require the thin film electrolyte with low electrical conductivity and high ionic conductivity. In addition, the thin-film electrolyte should be provided with high stability upon contacting with Li+ metal anode.
Although Li-sulfide and Li-oxide electrolytes have higher ionic conductivity (10−3∼10−4 S·cm−1), these two electrolytes are easily decomposed when being in contact with Li anode and the applied potential is up to 5.0 V.5,8) Lithium phosphorus oxynitride (LiPON) thin film, initially reported by Oak Ridge National Laboratory9) in 1990s, has gradually become a better alternative as the solid electrolyte. LiPON acronym describes a class of compounds with the general composition LixPOyNz, wherein the stoichiometric coefficients: x = 2y + 3z − 5.10) LiPONs can contact with Li anode and maintain stable chemical properties, nevertheless, they have relatively high Li-ion conductivity (10−6∼10−8 S·cm−1).
The key challenge to synthesize flexible LiPON thin film is to determine the appropriate substrate material and preparation method which may affect many different performances, such as capacity, cycling stability, adhesion together with stability. To the best of our knowledge, there are few relevant reports that the effect of substrates materials on the structure, morphology and electrochemical performance of LiPON thin films under the same preparation conditions. For the preparation method of LiPON thin films, there are many methods such as radio-frequency (RF) sputtering technique, atomic layer deposition, ion beam orienting assembly, electron beam evaporation, plasma-assisted direct vapor deposition, pulsed laser deposition and so on.11–13) Among these methods, the RF sputtering is widely used because it is simple, low-cost and MEMS-compatible, furthermore, it can provide higher ionic conductivity than other materials.
The aim of this study is to investigate the microstructure and electrochemical performance of LiPON films by employing different substrate materials and RF sputtering times. The characteristics and behaviors of films are studied systematically by using X-ray diffraction (XRD), atomic force microscopy (AFM) and electrochemical measurements. The results indicate that choosing appropriate substrate material or preparation method may be one key in the synthesis of LiPON films for advanced all-solid-state batteries.
The thin film batteries were prepared by RF magnetron sputtering method, which shown in Fig. 1 with (a) different substrates, (b) different sputtering times, respectively. As shown in Fig. 1(a), the LiPON films were deposited on the Cu, Mo, Al and FTO (SnO2: F) glass electrode layers using 50 W power, respectively, and then the Cu electrode layer was deposited at 100 W power on the LiPON solid electrolyte layer. Cylindrical Li3PO4 and Cu ceramic targets of 8 cm in diameter were used. No changes in target composition were observed with time and usage. The deposition chamber’s base pressure was 1.6 × 10−4 Pa, and during the deposition of Cu electrode layer, the gas pressure was maintained as a constant at 0.5 Pa. The substrate-to-target distance was 100 mm. Depositions of Li3PO4 and Cu layers were performed for 120 and 30 min, respectively. Finally, we fabricated the all-solid-state Li+ batteries with Cu, Mo, Al and FTO substrate, respectively. As shown in Fig. 1(b), the LiPON buffer layers were deposited on an FTO electrode layer with sputtering at 0.5, 1.0, 1.5 and 2.0 h. The preparation process is similar to the above method.
Schematic illustration of the all-solid-state lithium battery: (a) with different substrates materials; (b) with different sputtering times.
To investigate the crystallographic phases of the films, coupled θ–2θ XRD scans were performed in the range 2θ (20°–80°) by using the Cu Kα1 line of an X-ray source (Rigaku D/max2550). The surface morphologies of films were measured by AFM (DI Nanoscope IIIA Multimode). The values of resistivity were calculated from the sheet resistance which measured by a four point probe. All the electrochemical experiments were carried out with Mac Pile II system (Bio-Logic) in a dry atmosphere.
Figure 2(a) presents the XRD patterns of LiPON thin films which were deposited on different substrate materials. From the curve of FTO-sub film, the C-Li3PO4 polycrystalline phase occurs which shows six strong diffraction peaks of the orthorhombic phase. These peaks can be indexed to (021), (220), (131), (222), (331) and (260) of Li3PO4 crystal planes, respectively.6) For the Cu and Mo substrate, only two characteristic peaks Mo (110) and Cu (301) are found in the samples, respectively. In these two curves, the phenomenon of no other diffraction peak indicates that A-LiPON possesses amorphous structure. For the Al substrate, the crystalline quality of film is degraded. Moreover, no peak of metal phase can be observed which confirms that only amorphous structure appears. It can be interpreted by that the lattice constant of obtained films depends on the choice of substrate material. All curves suggest that the LiPON thin films are amorphous and independent of substrates. It is advantageous for battery applications because the ionic conductivity of amorphous films is generally higher than that of thin films containing crystal structure.
XRD patterns of LiPON films: (a) with different substrates; (b) with different sputtering times.
As shown in Fig. 2(b), the LiPON films were prepared at different sputtering time. For all samples, there exit six strong diffraction peaks corresponding to the orthorhombic phase. Combined with the above analysis, it can be proved that the films are amorphous from a side perspective.14–16) The different strengths of diffraction peaks indicate that the sputtering time could influence the diffusion and migration of atoms on the substrate. All these results elucidate that the LiPON film can be optimal synthesizes via controlling sputtering time.
As displayed in Fig. 3(a)–(d), the AFM images of the LiPON films on different substrates were measured with a scanning area of 1 µm. We can find that each LiPON film exhibits a rough surface with some protuberances which perhaps derived from phosphate crystals. According to related research, the rough surface may impede conduction efficiency of interfacial Li+ ions. Hence, it is clear that the film with rough surface is more suitable for application as a thin film Li+ ion battery electrolyte. From the details of this figure, the FTO-sub film has relatively rough comparing with other substrates, although the difference of flatness is not too obvious.17)
AFM images of LiPON films prepared on different substrate materials: (a) Al; (b) Cu; (c) Mo; (d) FTO.
Figure 4(a)–(d) presents the morphologies of LiPON films deposited at sputtering times of 0.5, 1.0, 1.5 and 2.0 h, respectively. As indicated in Fig. 4(a), the LiPON film with shorter time (0.5 h) possessed a high density of vacancy defects, which led to a large surface roughness. Figure 4(c–d) show the morphologies of films sputtering at the longer time (1.5 h and 2.0 h) which the surfaces become more rough. As we know, long deposition time may bring the grains to agglomerate with each other through a thermally activated growth mechanism.17,18)
AFM images of LiPON films which deposited on the FTO substrates prepared at different sputtering times: (a) 0.5 h; (b) 1.0 h; (c) 1.5 h; (d) 2.0 h.
Figure 5(a) and (b) illustrate the resistivity values calculated from the sheet resistance which measured by a four-point probe method. It should be mentioned that the smaller electronic conductivity is beneficial, because it can reduce the self-discharge rate of Li+ ion batteries. As shown in Fig. 5(a), LiPON films with Cu, Al, Mo and FTO electrodes exhibit the resistivity values at 0.34, 2.78, 3.32 and 6.01 Ω·cm, respectively, which demonstrate that the FTO-sub film may inhibit the negative effect of self-discharge efficiently comparing with other samples.14,19,20) Figure 5(b) shows that the film at the highest time (2.0 h) holds the higher resistivity (8.67 Ω·cm) than other films which may suppress self-discharge efficiency.14,18,21,22) Further research will continue.
The resistivity of LiPON films: (a) deposited on different substrate materials; (b) deposited with different sputtering times.
All-solid-state thin-film Li+ batteries were fabricated by depositing LiPON with different substrates and sputtering times. The discharge curves in Fig. 6 indicate that the curves shift downward to a lower voltage which may be caused by the resistance of batteries. As presented in Fig. 6(a), the LiPON battery on FTO substrate exhibits the highest discharge capacity of about 24 mAh, so it has a strong appeal for developing large capacity Li+ batteries. Nevertheless, the rapid drop in battery voltage would be avoided by improving the interface contact of the electrode film.18,23) Figure 6(b) illustrates that the LiPON battery at sputtering times (1.5 h) can exhibit the highest discharge capacity of about 23.5 mAh. It should be explained that the residual stress are affected by sputtering time during the LiPON deposition process, and the largest value appears at 1.5 h sputtering time. This residual stress (1.5 h) is so large that it can prevent the formation of polycrystalline LiPON structure, which benefits for the increase of capacity in the all-solid-state Li+ batteries.2,18,23,24)
Discharge curves of all-solid-state thin-film lithium battery: (a) with different substrate materials; (b) with different sputtering times.
In conclusion, we have manifested that choosing the appropriate substrate material and preparation method are vital for excellent performance of LiPON films. For the preparation with different substrates, the results indicate that the small particles of FTO-sub LiPON are aggregated together and form the particles clusters, which contribute to a relatively rough surface and high discharge capacities of the lithium cells during the cycle test. For the preparation with different RF sputtering times, the results reveal that long sputtering time (1.5 and 2.0 h) could prevent the formation of phases, which benefit for the increase of capacity in the all-solid-state Li+ batteries.
This work was funded by the National Natural Science Foundation of China (No. 11375112) and Science and Technology Commission of Shanghai (No. 16010500500, 15520500200).
