- Email: [email protected]
Investigating continuous co-intercalation of solvated lithium ions and graphite exfoliation in propylene carbonate-based electrolyte solutions
Journal of Power Sources 373 (2018) 110–118
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Investigating continuous co-intercalation of solvated lithium ions and graphite exfoliation in propylene carbonate-based electrolyte solutions
MARK
Hee-Youb Song, Soon-Ki Jeong∗ Department of Chemical Engineering, Soonchunhyang University, Asan, Chungnam 336-745, Republic of Korea
H I G H L I G H T S exfoliation is a problem in propylene carbonate (PC)-based electrolytes. • Graphite reactions affecting SEI formation were studied by in-situ AFM and CV. • Interfacial changes were observed over 10 cycles in ethylene carbonate-based electrolytes. • No blisters caused graphite exfoliation in PC-based electrolytes. • Overlapping • PC-solvated Li ions can pass through the blister structures freely.
A R T I C L E I N F O
A B S T R A C T
Keywords: Lithium ion batteries Graphite exfoliation Propylene carbonate Solid electrolyte interphase Co-intercalation reaction In situ atomic force microscopy
Forming an effective solid electrolyte interphase (SEI) is a significant issue in lithium ion batteries that utilize graphite as a negative electrode material, because the SEI determines the reversibility of the intercalation and de-intercalation of lithium ions into graphite for secondary batteries. In propylene carbonate (PC)-based electrolyte solutions, ceaseless co-intercalation of solvated lithium ions takes place because no effective SEI is formed. It is indisputable that this continuous co-intercalation leads to graphite exfoliation; however, the reason for this is currently not well understood. In this study, we investigate interfacial reactions that contribute to SEI formation on highly oriented pyrolytic graphite (HOPG) in ethylene carbonate (EC) and PC-based electrolyte solutions by in situ atomic force microscopy. The blisters formed on HOPG after the decomposition of solvated lithium ions within the graphite layers do not change over the course of ten electrochemical cycles in an ECbased electrolyte solution. In contrast, when cycling in PC-based electrolytes, the blisters continually change, and the height at the vicinity of the graphite edge plane increases. These morphological changes are attributed to the continuous co-intercalation of solvated lithium ions in PC-based electrolyte solutions.
1. Introduction In electrochemical systems, electrolyte solutions generally act as paths that allow ions to flow between the positive and negative electrodes during oxidation and reduction reactions. However, they also play another significant role in lithium ion batteries (LIBs). Commercial secondary LIBs have a high operating voltage because graphite is used as the negative electrode, where the intercalation and de-intercalation of lithium ions take place at extremely low potentials (0.25–0.0 V vs. Li+/Li) during charge and discharge processes [1–6]. For this reason, even organic electrolyte solutions with wider potential windows than aqueous electrolyte solutions will reductively decompose before the intercalation of lithium ions into graphite negative electrodes can take place [7–9]. In addition, it is easier to intercalate solvated lithium ions
∗
into graphite because of lower activation energy than that of lithium ion intercalation [10–12]. If these reactions occur continually, graphite negative electrodes never allow the intercalation of lithium ions during the charge process. Therefore, it is important to suppress further decomposition of the electrolyte solution and co-intercalation of solvated lithium ions to ensure reversible intercalation and de-intercalation of lithium ions at graphite negative electrodes in LIBs. Surface films commonly form within/on graphite negative electrodes during the reductive decomposition of organic electrolyte solutions [5–9]. These surface films determine the reversibility of the graphite negative electrodes as they are able to suppress further decomposition of the electrolyte solution and co-intercalation of solvated lithium ions by forming an effective surface film called a solid electrolyte interphase (SEI) [13,14]. An effective SEI will only conduct
Corresponding author. E-mail address: [email protected] (S.-K. Jeong).
http://dx.doi.org/10.1016/j.jpowsour.2017.11.015 Received 15 July 2017; Received in revised form 30 September 2017; Accepted 4 November 2017 0378-7753/ © 2017 Published by Elsevier B.V.
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
lithium ions, not electrons. Hence, the formation of an effective SEI is considered to be a significant issue at graphite negative electrodes. Thus, electrolyte solutions must not only act as ion conduction pathways, but also as SEI-forming reagents in order to ensure reversible intercalation and de-intercalation of lithium ions at the graphite negative electrodes in LIBs. It is well known that the properties of the SEI are greatly dependent on the type of organic solvent used [15,16]. Ethylene carbonate (EC)based solutions have been most widely used as electrolytes in LIBs owing to its outstanding SEI-forming abilities [7–9,17–19]. On the other hand, continuous co-intercalation and decomposition reactions of solvated lithium ions take place without the formation of effective SEI in propylene carbonate (PC)-based electrolyte solutions [15,16,19]. Consequently, graphite negative electrodes show good reversibility (lithium intercalation and de-intercalation) in EC-based electrolyte solutions but irreversibility (graphite exfoliation) in PC-based electrolyte solutions. However, EC is also co-intercalated with lithium ions into graphite negative electrodes at about 1 V vs. Li+/Li before the intercalation of unsolvated lithium ions, in the same manner as the co-intercalation of PC-solvated lithium ions. After this, reductive decomposition products are formed within the graphite layers, but there are significant differences between the subsequent reactions in EC- and PCbased electrolyte solutions, as mentioned above. This indicates that the decomposition products within the graphite layers play a significant role in suppressing further co-intercalation of solvated lithium ions. Indeed, surface decomposition reaction that form surface film on graphite take place after the suppression of the co-intercalation reaction [18,20]. Therefore, the co-intercalation of solvated lithium ions must be suppressed first in order to allow completion of SEI formation [21–23]. SEI at graphite negative electrodes have been investigated by various in situ techniques, including in situ atomic force microscopy (AFM), and in situ Fourier transform infrared spectroscopy (FTIR), which are powerful tools for investigating the SEI formation process and decomposition products [16–18,24,25]. Additionally, the reason for graphite exfoliation in PC-based electrolyte solutions has been explained using steric hindrance and gas evolution models [16,19,26–29]. There is no doubt that continuous co-intercalation reactions cause graphite exfoliation. However, the reason for continuous co-intercalation reactions at graphite negative electrodes in PC-based electrolyte solutions is not clearly understood. In this study, we used fast cyclic voltammetry (CV) scan rates of 5 and 10 mV s−1 and investigated the SEI formation process via in situ AFM to understand interfacial reactions between graphite and PC-based electrolyte solutions. In addition, intercalation and de-intercalation reactions into/from graphite layers were investigated using in situ Raman spectroscopy of graphite during oxidation and reduction in a PC-based electrolyte solution.
Fig. 1. Cyclic voltammograms rate = 0.5 mV s−1.
of
HOPG
in
1
mol
dm−3
LiClO4/PC.
Scan
observations were performed in an Ar-filled glove box. In situ Raman spectroscopy was performed to investigate the intercalation and de-intercalation of solvated lithium ions at the graphite negative electrode in 1 mol dm−3 LiClO4/PC. A three-electrode cell was assembled using a quartz cell in an Ar-filled glove box [4]. HOPG and lithium foil were used as the working electrode and the reference/ counter electrodes, respectively. The electrochemical cell was sealed to avoid exposure to air during the in situ Raman measurements. An argon ion laser (514.5 nm line) was scattered on HOPG during CV measurement at 5 mV s−1. Raman spectra were obtained using a triple monochromator (Jobin-Yvon, T64000) equipped with a charge-coupled device detector. 3. Results and discussion 3.1. Effect of CV scan rate on electrochemical properties Fig. 1 shows cyclic voltammograms of HOPG in 1 mol dm−3 LiClO4/ PC at a low scan rate (0.5 mV s−1). A cathodic current was confirmed below 1 V, which corresponds to the co-intercalation of solvated lithium ions into graphite layers. All the solvated lithium ions within the graphite were reductively decomposed during the reduction process, as no anodic peak can be seen. Furthermore, the cathodic current increased with increasing cycle number, which can be attributed to graphite exfoliation increasing the surface area. However, both reduction and oxidation peaks could be seen in 1 mol dm−3 LiClO4/PC at a fast scan rate (10 mV s−1), as shown in Fig. 2a. Although reduction peaks were still observed between 1 and 0.5 V during each of the five cycles, the cathodic current decreased after the first cycle, contrary to the result at the low scan rate. Moreover, a small anodic current was observed that increased in intensity after the first cycle. The results indicated that at the fast scan rate in PC-based electrolyte solutions, reversible reduction and oxidation reactions took place above the lithium ion intercalation/de-intercalation potential. However, on returning to the lower scan rate of 0.5 mV s−1 after the five cycles at the fast scan rate, only the cathodic current indicating an irreversible reaction was observed, with the oxidation peak having disappeared (Fig. 2b). It was founded that electrochemical properties are affected by the CV scan rate, and that graphite exfoliation in PC-based electrolyte solutions can be suppressed for five cycles at a fast scan rate without forming an effective SEI. Fig. 2c shows cyclic voltammograms of HOPG in 1 mol dm−3 LiClO4/EC + DEC. During the first cycle, three reduction peaks were observed below 1 V, with an additional peak at about 0 V that was attributed to the intercalation of lithium ions. These reduction peaks disappeared after the first cycle. Based on the literature, the
2. Experimental Electrochemical properties were investigated by CV in either a 1:1 (by volume) mixture of EC and diethyl carbonate (DEC) (EC + DEC) or PC, with both solutions containing 1 mol dm−3 LiClO4. CV was carried out to investigate the effect of scan rate at 0.5 and 10 mV s−1, respectively, with a laboratory-made three-electrode cell [30]. Highly oriented pyrolytic graphite (HOPG) was used as a working electrode. Lithium foil was used for reference and counter electrodes. A specially fabricated three-electrode cell was used for in situ AFM [31]. HOPG was used as a model electrode that allowed nanoscale changes to be observed due to its flat surface. Morphological changes on the HOPG basal plane were observed during CV at fast scan rates (5 and 10 mV s−1). Lithium foil was used for reference and counter electrodes. EC + DEC or PC containing 1 mol dm−3 LiClO4 were selected as electrolyte solutions to allow a comparison of their different SEI formation processes. AFM observations were carried out in contact mode using a microscope (Molecular Imaging, PicoSPM) equipped with a potentiostat (Molecular Imaging, PicoStat). All in situ AFM 111
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
Fig. 2. Cyclic voltammograms of HOPG in (a, b) 1 mol dm−3 LiClO4/PC and (c, d) EC + DEC (1:1, vol. ratio) at 0.5 and 10 mV s−1. The scan rate was changed from 10 to 0.5 mV s−1 after 5 cycles.
3.2. Morphological changes in the HOPG basal plane
peaks at about 1, 0.75, and 0.5 V were related to co-intercalation and decomposition reactions within/on graphite, respectively [18,20]. The resulting decomposition products act as an effective SEI within/on graphite anodes, suppressing further co-intercalation and decomposition reactions in EC-based electrolyte solutions. In this case, an effective SEI was formed on the graphite anodes after the first cycle in spite of the fast scan rate, as shown by the lack of reduction peaks between 1 and 0.5 V after the first cycle. As a result, reversible intercalation and de-intercalation of lithium ion took place at the graphite anodes at low scan rates in EC-based electrolyte solutions (Fig. 2d). In 1 mol dm−3 LiClO4/PC, it was considered that the continuous cointercalation of solvated lithium ions into the graphite anode proceeded without the formation of an effective SEI due to the appearance of reduction peaks during five cycles at the fast scan rate that were comparable with those seen in EC-based electrolyte solutions. However, the reason for the difference between the results obtained with the EC- and PC-based electrolyte solutions is still unclear. In addition, although graphite exfoliation was not observed during five cycles at the fast scan rate, it would likely occur after prolonged cycles, as it is clear that continuous co-intercalation reactions lead to graphite exfoliation in PCbased electrolyte solutions. Thus, it is necessary to investigate the interfacial reactions that correspond to SEI formation on graphite in EC/ PC-based electrolyte solutions before graphite exfoliation takes place. In the next section, we focus on the interfacial reactions at the graphite negative electrodes at fast scan rates, with morphological changes of HOPG in EC- and PC-based electrolyte solutions being investigated by in situ AFM.
Firstly, AFM observations were carried out to investigate morphological changes in the HOPG basal plane during ten cycles. Fig. 3 shows AFM images of the HOPG basal plane initially and after one, two, and ten cycles in 1 mol dm−3 LiClO4/EC + DEC. Flat terrace and step structures for the basal and edge planes were observed before cycling (Fig. 3a). The height of the step was estimated to be ∼10 nm. After the first cycle, precipitates caused by the decomposition of the electrolyte solution were formed on the HOPG surface; however, this is not the focus of this study, where we focused on structural changes within the graphite layers instead of on the surface film. However, morphological changes of HOPG underneath precipitates cannot be observed clearly because of the existing precipitates. Therefore, we performed repeated AFM scans to remove the precipitates on the HOPG surface after the first cycle [18,22,23]. Many blisters were observed on the HOPG basal plane after the first cycle, and typical blister structures are denoted by arrows in Fig. 3b. Generally, blister structures are observed after decomposition of solvated lithium ions within graphite layers through cointercalation reactions during the initial reduction process [18,32]. In this case, co-intercalation of solvated lithium ions into graphite layers took place during SEI formation. Furthermore, the blister structures were unchanged during ten cycles, indicating that they existed stably within the graphite layers (Fig. 3c and d). It can be considered that such blister structures suppressed further co-intercalation and decomposition reactions into/within the graphite layers, acting as an effective SEI within the graphite layers in EC-based electrolyte solutions. Morphological changes in the HOPG basal plane after each cycle in 1 mol dm−3 LiClO4/PC are shown as 3D images in Fig. 4. The relevant 112
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
Fig. 3. AFM images of the HOPG basal plane (5 μm × 5 μm) after a given number of cycles in 1 mol dm−3 LiClO4/EC + DEC. CV scan rate = 10 mV s−1.
After ten cycles in 1 mol dm−3 LiClO4/PC, large blister structures were observed at the vicinity of the edge plane that were estimated to be about 160 nm high, as shown Fig. 5a. The AFM image was obtained over an extended area to allow the thickness of the surface film on the HOPG to be investigated. The thickness of the surface film on HOPG can be estimated by removing the surface film via repeated scanning of an area (indicated by the square on the AFM image). However, no decomposition products were observed on the HOPG surface, indicating that only electrode reactions took place within the graphite layers during ten cycles in 1 mol dm−3 LiClO4/PC. On the other hand, a surface film was observed on HOPG after ten cycles in 1 mol dm−3 LiClO4/EC + DEC that was estimated to be about 40 nm thick, as shown in Fig. 5b. These results suggest that co-intercalation of solvated lithium ions must be suppressed to allow formation of a surface film on graphite negative electrodes.
2D images are inset in the lower left corner of each 3D image. In this case, precipitates resulting from decomposition of electrolyte solution were not observed on the HOPG surface, meaning that repeated AFM scans were not needed after the first cycle. Blister structures indicating decomposition reactions within the graphite layers were observed on the HOPG basal plane after the first cycle in 1 mol dm−3 LiClO4/PC (Fig. 4b). However, morphological changes along with increasing surface height of the HOPG basal plane were confirmed during ten cycles, in contrast with what was seen in EC-based electrolyte solutions (Fig. 4b–f). These results indicated that the blister structures did not suppress further co-intercalation of solvated lithium ions in PC-based electrolyte solutions. Consequently, fresh blisters were formed within the graphite layers upon cycling, leading to the expansion of the graphite layers. In addition, the HOPG edge plane where blisters were formed during the first cycle (indicated by the arrow in Fig. 4b) became gradually higher and wider during the ten cycles. In general, decomposition products derived from electrolyte solutions do not possess sufficient electron conductivity for further decomposition of electrolyte solutions to occur. In this case, however, decomposition reactions took place even in blister formed within graphite layers in PC-based electrolyte solutions. The reason for graphite exfoliation is suggested to be attributable to the overlapping formation of blisters near the graphite edge plane in PC-based electrolyte solutions.
3.3. Co-intercalation of solvated lithium ions at fast CV scan rate in PCbased electrolyte solutions In situ AFM was carried out to investigate co-intercalation of solvated lithium ions at fast scan rates in PC-based electrolyte solutions. Fig. 6 shows cyclic voltammograms during three cycles and AFM images of HOPG during the first reduction process in 1 mol dm−3 113
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
Fig. 4. 3D AFM images of the HOPG basal plane (5 μm × 5 μm) after a given number of cycles in LiClO4/PC. CV scan 1 mol dm−3 rate = 10 mV s−1.
Fig. 5. AFM images of the HOPG basal plane (10 μm × 10 μm) after 10 cycles in (a) PC and (b) EC + DEC containing 1 mol dm−3 LiClO4.
114
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
Fig. 6. (a) Cyclic voltammogram and in situ AFM images of the HOPG basal plane (5 μm × 5 μm) at (b) 3.20–2.86 V, (c) 1.14–0.78 V, (d) 0.78–0.44 V, and (e) 0.44–0.09 V during reduction in 1 mol dm−3 LiClO4/PC. Scan rate = 5 mV s−1.
LiClO4/PC. The potential was swept at 5 mV s−1 to allow morphological changes to be observed in more detail. In this case, a similar cyclic voltammogram to that obtained at 10 mV s−1 (see Fig. 2a) was obtained; thus, the in situ AFM results should be suitable for the investigation of morphological changes in HOPG at fast scan rates. Three reduction peaks during the first reduction process were confirmed between 1 and 0.0 V, as shown Fig. 6a. Furthermore, a small anodic current was also observed above 1 V during oxidation, which increased in intensity after the first cycle. Significant morphological changes to HOPG were observed below 1 V. Each AFM image in Fig. 6 corresponds to a cathodic peak at a different potential. Hill structures were confirmed at about 1 V, indicating that PC was co-intercalated into the
graphite layers with lithium ions (Fig. 6c). Furthermore, the decomposition of the solvated lithium ions within the graphite was confirmed to occur because of blister formation below 0.78 V, but the decomposition products like precipitates on the HOPG basal plane were not observed until reduction had completed (Fig. 6d and e). Fig. 7 shows morphological changes in the HOPG basal plane during oxidation in 1 mol dm−3 LiClO4/PC. Interestingly, the blister structures that formed during the first reduction process continuously changed during the oxidation process, as indicated by the arrow (Fig. 7a–e). Moreover, the height of the HOPG surface decreased slightly below 2.70 V, which can be seen as the dark areas in Fig. 7f. This result indicates that the blister structures had vanished and suggests that not all
115
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
Fig. 7. In situ AFM images of the HOPG basal plane (5 μm × 5 μm) at (a) 0.09–0.27 V, (b) 0.27–0.62 V, (c) 1.32–1.66 V, (d) 1.66–2.00 V, (e) 2.35–2.70 V, and (f) 2.70–3.05 V during oxidation in 1 mol dm−3 LiClO4/PC. Scan rate = 5 mV s−1.
oxidation PC-solvated lithium ions could easily passes through the blisters within the graphite layers resulting from the de-intercalation of solvated lithium ions without accommodation in the graphite layers. Thus, PC-derived blister structures allow further intercalation and deintercalation of PC-solvated lithium ions, meaning that the blisters can conduct PC-solvated lithium ions. In general, an effective SEI should suppress further co-intercalation and decomposition reactions of solvated lithium ions at graphite negative electrodes to allow lithium intercalation as mentioned above. Hence, the blisters caused by the decomposition of PC-solvated lithium ions cannot act as an effective SEI because they allow continuous co-intercalation and decomposition reactions. Until now, it has been accepted that graphite exfoliation occurs at the same time as co-intercalation and decomposition reactions, hence, fresh graphite edge planes are exposed to PC-based electrolyte solutions which leads to continuous co-intercalation reactions [16,19,26–29]. However, this explanation needs to be revised in light of the evidence above, which suggests that PC-derived blister structures
of the solvated lithium ions had decomposed within the graphite layers during the first reduction process. Some of the solvated lithium ions existed stably within the graphite layers, and these undecomposed solvated lithium ions were de-intercalated from the graphite negative electrode during the oxidation process. Thus, the oxidation peak observed in PC-based electrolyte solutions at fast scan rates was due to the de-intercalation of solvated lithium ions. This result can be explained by the fact that the concentration gradient becomes precipitous at the fast scan rate because solvated lithium ions are instantly consumed at the interface between the graphite and the PC-based electrolyte solution. Thus, the co-intercalation reaction rate becomes faster than the decomposition reaction rate within the graphite layers at the fast scan rate. However, at the low scan rate the co-intercalation rate might be relatively slow compared with the decomposition rate within the graphite layers, as all of the intercalated solvated lithium ions decomposed during the reduction process. Only cathodic peaks could be observed at the low scan rate in PC-based electrolyte solutions. In addition, during 116
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
Fig. 8. (a) Cyclic voltammogram and in situ Raman spectra of HOPG during (b) reduction and (c) oxidation in 1 mol dm−3 LiClO4/PC. Scan rate = 5 mV s−1.
solutions. In addition, de-intercalation of solvated lithium ions from HOPG was confirmed during oxidation at fast scan rates, suggesting that PC-solvated lithium ions can pass through the blister structures freely. This explains the continuous co-intercalation of solvated lithium ions in PC-based electrolyte solutions. The decomposition reactions within graphite layers were mitigated at fast scan rates, suggesting that they can be controlled by using various charge currents. Therefore, it is expected that the irreversible capacity degradation of LIBs in PC-based solutions can be reduced at high currents.
within graphite layers allow solvated lithium ions to pass through and intercalate, leading to continuous co-intercalation reactions. Furthermore, overlapping formation of blisters near the graphite edge plane causes graphite exfoliation in PC-based electrolyte solutions. Fig. 8 shows in situ Raman spectra of HOPG in 1 mol dm−3 LiClO4/ PC during the first cycle at 5 mV s−1. In this case, both reduction and oxidation peaks could be clearly observed because both the edge and basal planes of HOPG were used for reduction and oxidation reactions (Fig. 8a). Only the E2g2 (interior)-band at 1582 cm−1 corresponding to no intercalation within the graphite layers was observed above 1.2 V, while the E2g2 (boundary)-band at 1600 cm−1 corresponding to the presence of intercalates within the graphite layers appeared below 1.2 V in the Raman spectra (Fig. 8b). This result shows that co-intercalation of solvated lithium ions into HOPG took place below 1.2 V. The E2g2 (boundary)-band increased in intensity between 1.2 and 0.0 V, which was attributed to the continuous co-intercalation of solvated lithium ions. Additionally, the shift to original position before intercalation reaction of the E2g2 (boundary)-band during oxidation was due to the de-intercalation of solvated lithium ions from the graphite layers. This result strongly supports the in situ AFM results described above.
Acknowledgments This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (No. NRF-2017R1A2B4010544). This work was supported by the Soonchunhyang University Research Fund. References [1] [2] [3] [4]
4. Conclusions
[5] [6] [7] [8] [9] [10]
In this paper, we investigated the reasons for continuous co-intercalation and graphite exfoliation in PC-based electrolyte solutions by in situ AFM. It was found that upon cycling, the height of the blister structures increased near to the HOPG edge plane, in spite of the presence of other blisters formed during the first cycle. The overlapping formation of blisters caused graphite exfoliation in PC-based electrolyte
[11]
117
Z. Ogumi, M. Inaba, Bull. Chem. Soc. Jpn. 71 (1998) 521–534. S. Flandrois, B. Simon, Carbon 37 (1999) 165–180. T. Ohzuku, Y. Iwakoshi, K. Sawai, J. Electrochem. Soc. 140 (1993) 2490–2498. M. Inaba, H. Yoshida, Z. Ogumi, T. Abe, M. Mizutani, M. Asano, J. Electrochem. Soc. 142 (1995) 20–26. D. Aurbach, Y. Ein-Eli, J. Electrochem. Soc. 142 (1995) 1746–1752. M. Winter, J.O. Besenhard, M.E. Spahr, P. Novak, Adv. Mater 10 (1998) 725–763. K. Xu, Chem. Rev. 104 (2004) 4303–4417. S.S. Zhang, J. Power Sources 162 (2006) 1379–1394. D. Aurbach, J. Power Sources 89 (2000) 206–218. T. Abe, H. Fukuda, Y. Iriyama, Z. Ogumi, J. Electrochem. Soc. 151 (2004) A1120–A1123. T. Abe, F. Sagane, M. Ohtsuka, Y. Iriyama, Z. Ogumi, J. Electrochem. Soc. 152
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
(2003) A257–A261. [22] S.-K. Jeong, M. Inaba, Y. Iriyama, T. Abe, Z. Ogumi, J. Power Sources 175 (2008) 540–546. [23] H.-Y. Song, T. Fukutsuka, K. Miyazaki, T. Abe, J. Electrochem. Soc. 163 (2016) A1265–A1269. [24] D. Aurbach, M.L. Daroux, P.W. Faguy, E. Yeager, J. Electrochem. Soc. 134 (1987) 1611–1620. [25] R. Imhof, P. Novak, J. Electrochem. Soc. 145 (1998) 1087–1087. [26] D. Aurbach, M. Koltypin, H. Teller, Langmuir 18 (2002) 9000–9009. [27] D. Aurbach, E. Zinigrad, Y. Cohen, H. Teller, Solid State Ionics 148 (2002) 405–416. [28] D. Aurbach, H. Teller, M. Koltypin, E. Levi, J. Power Sources 119–121 (2003) 2–7. [29] H.-L. Zhang, C.-H. Sun, F. Li, C. Liu, J. Tan, H.-M. Cheng, J. Phys. Chem. C 111 (2007) 4740–4748. [30] K.-S. Park, S.-K. Jeong, Y.–S. Kim, Appl. Mech. Mater. 835 (2016) 126–130. [31] Y.-S. Kim, S.-K. Jeong, J. Spectros. 2015 (2015) 1–6. [32] M. Inaba, Z. Siroma, A. Funabiki, Z. Ogumi, Langmuir 12 (1996) 1535–1540.
(2005) A2151–A2154. [12] Z. Ogumi, T. Abe, T. Fukutsuka, S. Yamate, Y. Iriyama, J. Power Source 127 (2004) 72–75. [13] E. Peled, J. Electrochem. Soc. 126 (1979) 2047–2051. [14] P. Verma, P. Maire, P. Novak, Electrochim. Acta 55 (2010) 6332–6341. [15] K. Xu, J. Electrochem. Soc. 156 (2009) A751–A755. [16] M. Inaba, Z. Siroma, Y. Kawatate, A. Funabiki, Z. Ogumi, J. Power Sources 68 (1997) 221–226. [17] S.-K. Jeong, M. Inaba, Y. Iriyama, T. Abe, Z. Ogumi, Electrochim. Acta 47 (2002) 1975–1982. [18] S.-K. Jeong, M. Inaba, T. Abe, Z. Ogumi, J. Electrochem. Soc. 148 (2001) A989–A993. [19] D. Aurbach, B. Markovsky, I. Weissman, E. Levi, Y. Ein-Eli, Electrochim. Acta 45 (1999) 67–86. [20] J.O. Besenhard, M. Winter, J. Wang, W. Biberacher, J. Power Sources 54 (1995) 228–231. [21] T. Abe, N. Kawabata, Y. Mizutani, M. Inaba, Z. Ogumi, J. Electrochem. Soc. 150
118
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Investigating continuous co-intercalation of solvated lithium ions and graphite exfoliation in propylene carbonate-based electrolyte solutions
MARK
Hee-Youb Song, Soon-Ki Jeong∗ Department of Chemical Engineering, Soonchunhyang University, Asan, Chungnam 336-745, Republic of Korea
H I G H L I G H T S exfoliation is a problem in propylene carbonate (PC)-based electrolytes. • Graphite reactions affecting SEI formation were studied by in-situ AFM and CV. • Interfacial changes were observed over 10 cycles in ethylene carbonate-based electrolytes. • No blisters caused graphite exfoliation in PC-based electrolytes. • Overlapping • PC-solvated Li ions can pass through the blister structures freely.
A R T I C L E I N F O
A B S T R A C T
Keywords: Lithium ion batteries Graphite exfoliation Propylene carbonate Solid electrolyte interphase Co-intercalation reaction In situ atomic force microscopy
Forming an effective solid electrolyte interphase (SEI) is a significant issue in lithium ion batteries that utilize graphite as a negative electrode material, because the SEI determines the reversibility of the intercalation and de-intercalation of lithium ions into graphite for secondary batteries. In propylene carbonate (PC)-based electrolyte solutions, ceaseless co-intercalation of solvated lithium ions takes place because no effective SEI is formed. It is indisputable that this continuous co-intercalation leads to graphite exfoliation; however, the reason for this is currently not well understood. In this study, we investigate interfacial reactions that contribute to SEI formation on highly oriented pyrolytic graphite (HOPG) in ethylene carbonate (EC) and PC-based electrolyte solutions by in situ atomic force microscopy. The blisters formed on HOPG after the decomposition of solvated lithium ions within the graphite layers do not change over the course of ten electrochemical cycles in an ECbased electrolyte solution. In contrast, when cycling in PC-based electrolytes, the blisters continually change, and the height at the vicinity of the graphite edge plane increases. These morphological changes are attributed to the continuous co-intercalation of solvated lithium ions in PC-based electrolyte solutions.
1. Introduction In electrochemical systems, electrolyte solutions generally act as paths that allow ions to flow between the positive and negative electrodes during oxidation and reduction reactions. However, they also play another significant role in lithium ion batteries (LIBs). Commercial secondary LIBs have a high operating voltage because graphite is used as the negative electrode, where the intercalation and de-intercalation of lithium ions take place at extremely low potentials (0.25–0.0 V vs. Li+/Li) during charge and discharge processes [1–6]. For this reason, even organic electrolyte solutions with wider potential windows than aqueous electrolyte solutions will reductively decompose before the intercalation of lithium ions into graphite negative electrodes can take place [7–9]. In addition, it is easier to intercalate solvated lithium ions
∗
into graphite because of lower activation energy than that of lithium ion intercalation [10–12]. If these reactions occur continually, graphite negative electrodes never allow the intercalation of lithium ions during the charge process. Therefore, it is important to suppress further decomposition of the electrolyte solution and co-intercalation of solvated lithium ions to ensure reversible intercalation and de-intercalation of lithium ions at graphite negative electrodes in LIBs. Surface films commonly form within/on graphite negative electrodes during the reductive decomposition of organic electrolyte solutions [5–9]. These surface films determine the reversibility of the graphite negative electrodes as they are able to suppress further decomposition of the electrolyte solution and co-intercalation of solvated lithium ions by forming an effective surface film called a solid electrolyte interphase (SEI) [13,14]. An effective SEI will only conduct
Corresponding author. E-mail address: [email protected] (S.-K. Jeong).
http://dx.doi.org/10.1016/j.jpowsour.2017.11.015 Received 15 July 2017; Received in revised form 30 September 2017; Accepted 4 November 2017 0378-7753/ © 2017 Published by Elsevier B.V.
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
lithium ions, not electrons. Hence, the formation of an effective SEI is considered to be a significant issue at graphite negative electrodes. Thus, electrolyte solutions must not only act as ion conduction pathways, but also as SEI-forming reagents in order to ensure reversible intercalation and de-intercalation of lithium ions at the graphite negative electrodes in LIBs. It is well known that the properties of the SEI are greatly dependent on the type of organic solvent used [15,16]. Ethylene carbonate (EC)based solutions have been most widely used as electrolytes in LIBs owing to its outstanding SEI-forming abilities [7–9,17–19]. On the other hand, continuous co-intercalation and decomposition reactions of solvated lithium ions take place without the formation of effective SEI in propylene carbonate (PC)-based electrolyte solutions [15,16,19]. Consequently, graphite negative electrodes show good reversibility (lithium intercalation and de-intercalation) in EC-based electrolyte solutions but irreversibility (graphite exfoliation) in PC-based electrolyte solutions. However, EC is also co-intercalated with lithium ions into graphite negative electrodes at about 1 V vs. Li+/Li before the intercalation of unsolvated lithium ions, in the same manner as the co-intercalation of PC-solvated lithium ions. After this, reductive decomposition products are formed within the graphite layers, but there are significant differences between the subsequent reactions in EC- and PCbased electrolyte solutions, as mentioned above. This indicates that the decomposition products within the graphite layers play a significant role in suppressing further co-intercalation of solvated lithium ions. Indeed, surface decomposition reaction that form surface film on graphite take place after the suppression of the co-intercalation reaction [18,20]. Therefore, the co-intercalation of solvated lithium ions must be suppressed first in order to allow completion of SEI formation [21–23]. SEI at graphite negative electrodes have been investigated by various in situ techniques, including in situ atomic force microscopy (AFM), and in situ Fourier transform infrared spectroscopy (FTIR), which are powerful tools for investigating the SEI formation process and decomposition products [16–18,24,25]. Additionally, the reason for graphite exfoliation in PC-based electrolyte solutions has been explained using steric hindrance and gas evolution models [16,19,26–29]. There is no doubt that continuous co-intercalation reactions cause graphite exfoliation. However, the reason for continuous co-intercalation reactions at graphite negative electrodes in PC-based electrolyte solutions is not clearly understood. In this study, we used fast cyclic voltammetry (CV) scan rates of 5 and 10 mV s−1 and investigated the SEI formation process via in situ AFM to understand interfacial reactions between graphite and PC-based electrolyte solutions. In addition, intercalation and de-intercalation reactions into/from graphite layers were investigated using in situ Raman spectroscopy of graphite during oxidation and reduction in a PC-based electrolyte solution.
Fig. 1. Cyclic voltammograms rate = 0.5 mV s−1.
of
HOPG
in
1
mol
dm−3
LiClO4/PC.
Scan
observations were performed in an Ar-filled glove box. In situ Raman spectroscopy was performed to investigate the intercalation and de-intercalation of solvated lithium ions at the graphite negative electrode in 1 mol dm−3 LiClO4/PC. A three-electrode cell was assembled using a quartz cell in an Ar-filled glove box [4]. HOPG and lithium foil were used as the working electrode and the reference/ counter electrodes, respectively. The electrochemical cell was sealed to avoid exposure to air during the in situ Raman measurements. An argon ion laser (514.5 nm line) was scattered on HOPG during CV measurement at 5 mV s−1. Raman spectra were obtained using a triple monochromator (Jobin-Yvon, T64000) equipped with a charge-coupled device detector. 3. Results and discussion 3.1. Effect of CV scan rate on electrochemical properties Fig. 1 shows cyclic voltammograms of HOPG in 1 mol dm−3 LiClO4/ PC at a low scan rate (0.5 mV s−1). A cathodic current was confirmed below 1 V, which corresponds to the co-intercalation of solvated lithium ions into graphite layers. All the solvated lithium ions within the graphite were reductively decomposed during the reduction process, as no anodic peak can be seen. Furthermore, the cathodic current increased with increasing cycle number, which can be attributed to graphite exfoliation increasing the surface area. However, both reduction and oxidation peaks could be seen in 1 mol dm−3 LiClO4/PC at a fast scan rate (10 mV s−1), as shown in Fig. 2a. Although reduction peaks were still observed between 1 and 0.5 V during each of the five cycles, the cathodic current decreased after the first cycle, contrary to the result at the low scan rate. Moreover, a small anodic current was observed that increased in intensity after the first cycle. The results indicated that at the fast scan rate in PC-based electrolyte solutions, reversible reduction and oxidation reactions took place above the lithium ion intercalation/de-intercalation potential. However, on returning to the lower scan rate of 0.5 mV s−1 after the five cycles at the fast scan rate, only the cathodic current indicating an irreversible reaction was observed, with the oxidation peak having disappeared (Fig. 2b). It was founded that electrochemical properties are affected by the CV scan rate, and that graphite exfoliation in PC-based electrolyte solutions can be suppressed for five cycles at a fast scan rate without forming an effective SEI. Fig. 2c shows cyclic voltammograms of HOPG in 1 mol dm−3 LiClO4/EC + DEC. During the first cycle, three reduction peaks were observed below 1 V, with an additional peak at about 0 V that was attributed to the intercalation of lithium ions. These reduction peaks disappeared after the first cycle. Based on the literature, the
2. Experimental Electrochemical properties were investigated by CV in either a 1:1 (by volume) mixture of EC and diethyl carbonate (DEC) (EC + DEC) or PC, with both solutions containing 1 mol dm−3 LiClO4. CV was carried out to investigate the effect of scan rate at 0.5 and 10 mV s−1, respectively, with a laboratory-made three-electrode cell [30]. Highly oriented pyrolytic graphite (HOPG) was used as a working electrode. Lithium foil was used for reference and counter electrodes. A specially fabricated three-electrode cell was used for in situ AFM [31]. HOPG was used as a model electrode that allowed nanoscale changes to be observed due to its flat surface. Morphological changes on the HOPG basal plane were observed during CV at fast scan rates (5 and 10 mV s−1). Lithium foil was used for reference and counter electrodes. EC + DEC or PC containing 1 mol dm−3 LiClO4 were selected as electrolyte solutions to allow a comparison of their different SEI formation processes. AFM observations were carried out in contact mode using a microscope (Molecular Imaging, PicoSPM) equipped with a potentiostat (Molecular Imaging, PicoStat). All in situ AFM 111
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
Fig. 2. Cyclic voltammograms of HOPG in (a, b) 1 mol dm−3 LiClO4/PC and (c, d) EC + DEC (1:1, vol. ratio) at 0.5 and 10 mV s−1. The scan rate was changed from 10 to 0.5 mV s−1 after 5 cycles.
3.2. Morphological changes in the HOPG basal plane
peaks at about 1, 0.75, and 0.5 V were related to co-intercalation and decomposition reactions within/on graphite, respectively [18,20]. The resulting decomposition products act as an effective SEI within/on graphite anodes, suppressing further co-intercalation and decomposition reactions in EC-based electrolyte solutions. In this case, an effective SEI was formed on the graphite anodes after the first cycle in spite of the fast scan rate, as shown by the lack of reduction peaks between 1 and 0.5 V after the first cycle. As a result, reversible intercalation and de-intercalation of lithium ion took place at the graphite anodes at low scan rates in EC-based electrolyte solutions (Fig. 2d). In 1 mol dm−3 LiClO4/PC, it was considered that the continuous cointercalation of solvated lithium ions into the graphite anode proceeded without the formation of an effective SEI due to the appearance of reduction peaks during five cycles at the fast scan rate that were comparable with those seen in EC-based electrolyte solutions. However, the reason for the difference between the results obtained with the EC- and PC-based electrolyte solutions is still unclear. In addition, although graphite exfoliation was not observed during five cycles at the fast scan rate, it would likely occur after prolonged cycles, as it is clear that continuous co-intercalation reactions lead to graphite exfoliation in PCbased electrolyte solutions. Thus, it is necessary to investigate the interfacial reactions that correspond to SEI formation on graphite in EC/ PC-based electrolyte solutions before graphite exfoliation takes place. In the next section, we focus on the interfacial reactions at the graphite negative electrodes at fast scan rates, with morphological changes of HOPG in EC- and PC-based electrolyte solutions being investigated by in situ AFM.
Firstly, AFM observations were carried out to investigate morphological changes in the HOPG basal plane during ten cycles. Fig. 3 shows AFM images of the HOPG basal plane initially and after one, two, and ten cycles in 1 mol dm−3 LiClO4/EC + DEC. Flat terrace and step structures for the basal and edge planes were observed before cycling (Fig. 3a). The height of the step was estimated to be ∼10 nm. After the first cycle, precipitates caused by the decomposition of the electrolyte solution were formed on the HOPG surface; however, this is not the focus of this study, where we focused on structural changes within the graphite layers instead of on the surface film. However, morphological changes of HOPG underneath precipitates cannot be observed clearly because of the existing precipitates. Therefore, we performed repeated AFM scans to remove the precipitates on the HOPG surface after the first cycle [18,22,23]. Many blisters were observed on the HOPG basal plane after the first cycle, and typical blister structures are denoted by arrows in Fig. 3b. Generally, blister structures are observed after decomposition of solvated lithium ions within graphite layers through cointercalation reactions during the initial reduction process [18,32]. In this case, co-intercalation of solvated lithium ions into graphite layers took place during SEI formation. Furthermore, the blister structures were unchanged during ten cycles, indicating that they existed stably within the graphite layers (Fig. 3c and d). It can be considered that such blister structures suppressed further co-intercalation and decomposition reactions into/within the graphite layers, acting as an effective SEI within the graphite layers in EC-based electrolyte solutions. Morphological changes in the HOPG basal plane after each cycle in 1 mol dm−3 LiClO4/PC are shown as 3D images in Fig. 4. The relevant 112
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
Fig. 3. AFM images of the HOPG basal plane (5 μm × 5 μm) after a given number of cycles in 1 mol dm−3 LiClO4/EC + DEC. CV scan rate = 10 mV s−1.
After ten cycles in 1 mol dm−3 LiClO4/PC, large blister structures were observed at the vicinity of the edge plane that were estimated to be about 160 nm high, as shown Fig. 5a. The AFM image was obtained over an extended area to allow the thickness of the surface film on the HOPG to be investigated. The thickness of the surface film on HOPG can be estimated by removing the surface film via repeated scanning of an area (indicated by the square on the AFM image). However, no decomposition products were observed on the HOPG surface, indicating that only electrode reactions took place within the graphite layers during ten cycles in 1 mol dm−3 LiClO4/PC. On the other hand, a surface film was observed on HOPG after ten cycles in 1 mol dm−3 LiClO4/EC + DEC that was estimated to be about 40 nm thick, as shown in Fig. 5b. These results suggest that co-intercalation of solvated lithium ions must be suppressed to allow formation of a surface film on graphite negative electrodes.
2D images are inset in the lower left corner of each 3D image. In this case, precipitates resulting from decomposition of electrolyte solution were not observed on the HOPG surface, meaning that repeated AFM scans were not needed after the first cycle. Blister structures indicating decomposition reactions within the graphite layers were observed on the HOPG basal plane after the first cycle in 1 mol dm−3 LiClO4/PC (Fig. 4b). However, morphological changes along with increasing surface height of the HOPG basal plane were confirmed during ten cycles, in contrast with what was seen in EC-based electrolyte solutions (Fig. 4b–f). These results indicated that the blister structures did not suppress further co-intercalation of solvated lithium ions in PC-based electrolyte solutions. Consequently, fresh blisters were formed within the graphite layers upon cycling, leading to the expansion of the graphite layers. In addition, the HOPG edge plane where blisters were formed during the first cycle (indicated by the arrow in Fig. 4b) became gradually higher and wider during the ten cycles. In general, decomposition products derived from electrolyte solutions do not possess sufficient electron conductivity for further decomposition of electrolyte solutions to occur. In this case, however, decomposition reactions took place even in blister formed within graphite layers in PC-based electrolyte solutions. The reason for graphite exfoliation is suggested to be attributable to the overlapping formation of blisters near the graphite edge plane in PC-based electrolyte solutions.
3.3. Co-intercalation of solvated lithium ions at fast CV scan rate in PCbased electrolyte solutions In situ AFM was carried out to investigate co-intercalation of solvated lithium ions at fast scan rates in PC-based electrolyte solutions. Fig. 6 shows cyclic voltammograms during three cycles and AFM images of HOPG during the first reduction process in 1 mol dm−3 113
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
Fig. 4. 3D AFM images of the HOPG basal plane (5 μm × 5 μm) after a given number of cycles in LiClO4/PC. CV scan 1 mol dm−3 rate = 10 mV s−1.
Fig. 5. AFM images of the HOPG basal plane (10 μm × 10 μm) after 10 cycles in (a) PC and (b) EC + DEC containing 1 mol dm−3 LiClO4.
114
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
Fig. 6. (a) Cyclic voltammogram and in situ AFM images of the HOPG basal plane (5 μm × 5 μm) at (b) 3.20–2.86 V, (c) 1.14–0.78 V, (d) 0.78–0.44 V, and (e) 0.44–0.09 V during reduction in 1 mol dm−3 LiClO4/PC. Scan rate = 5 mV s−1.
LiClO4/PC. The potential was swept at 5 mV s−1 to allow morphological changes to be observed in more detail. In this case, a similar cyclic voltammogram to that obtained at 10 mV s−1 (see Fig. 2a) was obtained; thus, the in situ AFM results should be suitable for the investigation of morphological changes in HOPG at fast scan rates. Three reduction peaks during the first reduction process were confirmed between 1 and 0.0 V, as shown Fig. 6a. Furthermore, a small anodic current was also observed above 1 V during oxidation, which increased in intensity after the first cycle. Significant morphological changes to HOPG were observed below 1 V. Each AFM image in Fig. 6 corresponds to a cathodic peak at a different potential. Hill structures were confirmed at about 1 V, indicating that PC was co-intercalated into the
graphite layers with lithium ions (Fig. 6c). Furthermore, the decomposition of the solvated lithium ions within the graphite was confirmed to occur because of blister formation below 0.78 V, but the decomposition products like precipitates on the HOPG basal plane were not observed until reduction had completed (Fig. 6d and e). Fig. 7 shows morphological changes in the HOPG basal plane during oxidation in 1 mol dm−3 LiClO4/PC. Interestingly, the blister structures that formed during the first reduction process continuously changed during the oxidation process, as indicated by the arrow (Fig. 7a–e). Moreover, the height of the HOPG surface decreased slightly below 2.70 V, which can be seen as the dark areas in Fig. 7f. This result indicates that the blister structures had vanished and suggests that not all
115
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
Fig. 7. In situ AFM images of the HOPG basal plane (5 μm × 5 μm) at (a) 0.09–0.27 V, (b) 0.27–0.62 V, (c) 1.32–1.66 V, (d) 1.66–2.00 V, (e) 2.35–2.70 V, and (f) 2.70–3.05 V during oxidation in 1 mol dm−3 LiClO4/PC. Scan rate = 5 mV s−1.
oxidation PC-solvated lithium ions could easily passes through the blisters within the graphite layers resulting from the de-intercalation of solvated lithium ions without accommodation in the graphite layers. Thus, PC-derived blister structures allow further intercalation and deintercalation of PC-solvated lithium ions, meaning that the blisters can conduct PC-solvated lithium ions. In general, an effective SEI should suppress further co-intercalation and decomposition reactions of solvated lithium ions at graphite negative electrodes to allow lithium intercalation as mentioned above. Hence, the blisters caused by the decomposition of PC-solvated lithium ions cannot act as an effective SEI because they allow continuous co-intercalation and decomposition reactions. Until now, it has been accepted that graphite exfoliation occurs at the same time as co-intercalation and decomposition reactions, hence, fresh graphite edge planes are exposed to PC-based electrolyte solutions which leads to continuous co-intercalation reactions [16,19,26–29]. However, this explanation needs to be revised in light of the evidence above, which suggests that PC-derived blister structures
of the solvated lithium ions had decomposed within the graphite layers during the first reduction process. Some of the solvated lithium ions existed stably within the graphite layers, and these undecomposed solvated lithium ions were de-intercalated from the graphite negative electrode during the oxidation process. Thus, the oxidation peak observed in PC-based electrolyte solutions at fast scan rates was due to the de-intercalation of solvated lithium ions. This result can be explained by the fact that the concentration gradient becomes precipitous at the fast scan rate because solvated lithium ions are instantly consumed at the interface between the graphite and the PC-based electrolyte solution. Thus, the co-intercalation reaction rate becomes faster than the decomposition reaction rate within the graphite layers at the fast scan rate. However, at the low scan rate the co-intercalation rate might be relatively slow compared with the decomposition rate within the graphite layers, as all of the intercalated solvated lithium ions decomposed during the reduction process. Only cathodic peaks could be observed at the low scan rate in PC-based electrolyte solutions. In addition, during 116
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
Fig. 8. (a) Cyclic voltammogram and in situ Raman spectra of HOPG during (b) reduction and (c) oxidation in 1 mol dm−3 LiClO4/PC. Scan rate = 5 mV s−1.
solutions. In addition, de-intercalation of solvated lithium ions from HOPG was confirmed during oxidation at fast scan rates, suggesting that PC-solvated lithium ions can pass through the blister structures freely. This explains the continuous co-intercalation of solvated lithium ions in PC-based electrolyte solutions. The decomposition reactions within graphite layers were mitigated at fast scan rates, suggesting that they can be controlled by using various charge currents. Therefore, it is expected that the irreversible capacity degradation of LIBs in PC-based solutions can be reduced at high currents.
within graphite layers allow solvated lithium ions to pass through and intercalate, leading to continuous co-intercalation reactions. Furthermore, overlapping formation of blisters near the graphite edge plane causes graphite exfoliation in PC-based electrolyte solutions. Fig. 8 shows in situ Raman spectra of HOPG in 1 mol dm−3 LiClO4/ PC during the first cycle at 5 mV s−1. In this case, both reduction and oxidation peaks could be clearly observed because both the edge and basal planes of HOPG were used for reduction and oxidation reactions (Fig. 8a). Only the E2g2 (interior)-band at 1582 cm−1 corresponding to no intercalation within the graphite layers was observed above 1.2 V, while the E2g2 (boundary)-band at 1600 cm−1 corresponding to the presence of intercalates within the graphite layers appeared below 1.2 V in the Raman spectra (Fig. 8b). This result shows that co-intercalation of solvated lithium ions into HOPG took place below 1.2 V. The E2g2 (boundary)-band increased in intensity between 1.2 and 0.0 V, which was attributed to the continuous co-intercalation of solvated lithium ions. Additionally, the shift to original position before intercalation reaction of the E2g2 (boundary)-band during oxidation was due to the de-intercalation of solvated lithium ions from the graphite layers. This result strongly supports the in situ AFM results described above.
Acknowledgments This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (No. NRF-2017R1A2B4010544). This work was supported by the Soonchunhyang University Research Fund. References [1] [2] [3] [4]
4. Conclusions
[5] [6] [7] [8] [9] [10]
In this paper, we investigated the reasons for continuous co-intercalation and graphite exfoliation in PC-based electrolyte solutions by in situ AFM. It was found that upon cycling, the height of the blister structures increased near to the HOPG edge plane, in spite of the presence of other blisters formed during the first cycle. The overlapping formation of blisters caused graphite exfoliation in PC-based electrolyte
[11]
117
Z. Ogumi, M. Inaba, Bull. Chem. Soc. Jpn. 71 (1998) 521–534. S. Flandrois, B. Simon, Carbon 37 (1999) 165–180. T. Ohzuku, Y. Iwakoshi, K. Sawai, J. Electrochem. Soc. 140 (1993) 2490–2498. M. Inaba, H. Yoshida, Z. Ogumi, T. Abe, M. Mizutani, M. Asano, J. Electrochem. Soc. 142 (1995) 20–26. D. Aurbach, Y. Ein-Eli, J. Electrochem. Soc. 142 (1995) 1746–1752. M. Winter, J.O. Besenhard, M.E. Spahr, P. Novak, Adv. Mater 10 (1998) 725–763. K. Xu, Chem. Rev. 104 (2004) 4303–4417. S.S. Zhang, J. Power Sources 162 (2006) 1379–1394. D. Aurbach, J. Power Sources 89 (2000) 206–218. T. Abe, H. Fukuda, Y. Iriyama, Z. Ogumi, J. Electrochem. Soc. 151 (2004) A1120–A1123. T. Abe, F. Sagane, M. Ohtsuka, Y. Iriyama, Z. Ogumi, J. Electrochem. Soc. 152
Journal of Power Sources 373 (2018) 110–118
H.-Y. Song, S.-K. Jeong
(2003) A257–A261. [22] S.-K. Jeong, M. Inaba, Y. Iriyama, T. Abe, Z. Ogumi, J. Power Sources 175 (2008) 540–546. [23] H.-Y. Song, T. Fukutsuka, K. Miyazaki, T. Abe, J. Electrochem. Soc. 163 (2016) A1265–A1269. [24] D. Aurbach, M.L. Daroux, P.W. Faguy, E. Yeager, J. Electrochem. Soc. 134 (1987) 1611–1620. [25] R. Imhof, P. Novak, J. Electrochem. Soc. 145 (1998) 1087–1087. [26] D. Aurbach, M. Koltypin, H. Teller, Langmuir 18 (2002) 9000–9009. [27] D. Aurbach, E. Zinigrad, Y. Cohen, H. Teller, Solid State Ionics 148 (2002) 405–416. [28] D. Aurbach, H. Teller, M. Koltypin, E. Levi, J. Power Sources 119–121 (2003) 2–7. [29] H.-L. Zhang, C.-H. Sun, F. Li, C. Liu, J. Tan, H.-M. Cheng, J. Phys. Chem. C 111 (2007) 4740–4748. [30] K.-S. Park, S.-K. Jeong, Y.–S. Kim, Appl. Mech. Mater. 835 (2016) 126–130. [31] Y.-S. Kim, S.-K. Jeong, J. Spectros. 2015 (2015) 1–6. [32] M. Inaba, Z. Siroma, A. Funabiki, Z. Ogumi, Langmuir 12 (1996) 1535–1540.
(2005) A2151–A2154. [12] Z. Ogumi, T. Abe, T. Fukutsuka, S. Yamate, Y. Iriyama, J. Power Source 127 (2004) 72–75. [13] E. Peled, J. Electrochem. Soc. 126 (1979) 2047–2051. [14] P. Verma, P. Maire, P. Novak, Electrochim. Acta 55 (2010) 6332–6341. [15] K. Xu, J. Electrochem. Soc. 156 (2009) A751–A755. [16] M. Inaba, Z. Siroma, Y. Kawatate, A. Funabiki, Z. Ogumi, J. Power Sources 68 (1997) 221–226. [17] S.-K. Jeong, M. Inaba, Y. Iriyama, T. Abe, Z. Ogumi, Electrochim. Acta 47 (2002) 1975–1982. [18] S.-K. Jeong, M. Inaba, T. Abe, Z. Ogumi, J. Electrochem. Soc. 148 (2001) A989–A993. [19] D. Aurbach, B. Markovsky, I. Weissman, E. Levi, Y. Ein-Eli, Electrochim. Acta 45 (1999) 67–86. [20] J.O. Besenhard, M. Winter, J. Wang, W. Biberacher, J. Power Sources 54 (1995) 228–231. [21] T. Abe, N. Kawabata, Y. Mizutani, M. Inaba, Z. Ogumi, J. Electrochem. Soc. 150
118