Formation of thermally resistant films induced by vinylene carbonate additive on a hard carbon anode for lithium ion batteries at elevated temperature

Abstract

Vinylene carbonate (VC)-induced film formation in a LiFePO₄/hard carbon (HC) cell is clarified based on mass spectroscopic analysis. VC is found to induce the formation of different types of thermally resistant organophosphates on the HC surface for different solvents and their formation mechanisms are elucidated. Formation of the organophosphates is also found to contribute to the improved cycle performance at elevated temperature due to their thermally stable characteristics.

---

Introduction

Utilization of lithium ion batteries (LIBs) is currently shifting to large-scale electronic applications such as electric vehicles (EVs), hybrid electric vehicles (HEVs) as well as other stationary battery storage systems for home or industry use.^1–5 Many efforts have been made to improve the battery performance;^6–8 meanwhile, the environment for the use of LIBs becomes more rigorous because most of the large-scale electronic devices are used under a relatively high-temperature environment. In particular, daily use of the thermally instable LIBs would directly impact on our public safety. Development of thermally stable LIBs are thus more focused to match the severe requirement on the battery design with high performance and safety. Enhancement of the thermal stability of electrolyte is considered a direct approach to solve the aforementioned safety issue since the electrolyte is mainly composed of organic carbonate solvent and LiPF₆ salt, which tend to rapidly degrade at elevated temperature, reducing the battery performance and safety. This can be evidenced by a recent study, showing that degradation products, particularly for organophosphates, tend to be formed at elevated temperature, corresponding to the cell's cycle performance.⁹

Adding electrolyte additives in a battery has been reported effective in increasing the cell's thermal stability.¹⁰ Among the electrolyte additives, vinylene carbonate (VC) has been so far demonstrated as a promising one for this issue. Several cathodes^11–18 and anodes^14,17–23 were shown to have enhanced cell's cycle life by use of the VC additive. Due to its high reactivity, VC tends to induce polymerization in the electrolyte solution, leading to the formation of the passivating film on the anode surface.^20,21 The formed film was considered important in preventing the anode surface from overreacting with other chemical compounds in the cell, particularly at elevated temperature. In addition, in a LiFePO₄/graphite cell tested by Ouatani et al.,²⁴ the VC polymerization was observed on the anode side; while on the other hand, it was hardly identified for the cathode side. With the polymerization function, the properties of the passivating films formed on the anode surface were improved, resulting in better cell's cycle life. This provides a straightforward evidence for the fact that VC has been known as an effective film-forming additive for the LIBs composed of graphite anode and been most widely used in the practical manufacture. Furthermore, a detailed mechanism for the film formation induced by VC additive has been reported by Liu et al.,²⁵ based on their quantitative analysis via liquid chromatography mass spectroscopy (LC-MS) together with direct analysis in real time mass spectroscopy (DART-MS). This study takes a step forward in clarifying the reason of the improved cycle life assisted by the VC additive.

In addition to graphite, hard carbon (HC) is also able to accommodate lithium due to the significant porosity formed by the random arrangement of small-dimensional graphene layers.²⁶ More importantly, higher rate performance of HC, compared to graphite,^27–29 allows it to be a potential candidate for high power density application, such as EV. However, the interfacial phenomenon between the HC and electrolyte is still not sufficiently clear, in particular for the case at elevated temperature. Thus, some fundamental investigation relating the cell performance to the phenomena occurring on electrode (HC)/electrolyte interface is considered necessary. To extend our previous research, we study the effect of VC additive on HC anode at 60 °C. We not only analyse the formed products in the electrolyte after cycling by use of LC-MS but also specify the formed products on the electrode surface using DART-MS. Based on the analysis results, the film formation induced by VC on the HC surface is clarified, which, to our best knowledge, has not been reported.

Experimental section

Preparation of LiFePO₄/HC cell

The cathode material used was LiFePO₄ (Sumitomo Osaka Cement Co., Ltd.) powders, which were mixed with 2 wt% vapor growth carbon fiber (VGCF), 2 wt% acetylene black (AB) and 2 wt% activated carbon fiber (ACF), together with a 4 wt% polyvinylidenedifluoride (PVDF) binder to prepare the slurry. After well mixing, the slurry was then pasted on a carbon coated aluminum foil sheet, followed by a preheating procedure at 80 °C for 10 min. Through a subsequent roller-press and a drying treatment under vacuum at 160 °C for 6 h, the cathode was completed. As for the anode, its slurry was composed of 93 wt% HC (Sumitomo Bakelite Co., Ltd.) powders, 1 wt% VGCF, 2 wt% AB and 4 wt% water-based acrylic binder, which was subsequently coated on a copper based foil sheet. The following electrode-preparing procedures were similar to the cathode, as described above. In addition, the area capacity for the cathode and anode is ca. 1 and 2 mA h cm⁻², respectively, leading to a N/P ratio of 2.

Electrochemical evaluation and analysis

For the electrochemical tests, a coin-type test cell (CR2032) was used. The LiFePO₄ cathode and HC anode were assembled in the cell with 1 M LiPF₆ in various solvent mixtures, which are summarized in Table 1. A charge–discharge machine (BLS series, Keisokuki Center Co., Ltd.) was applied to galvanostatically perform the electrochemical tests at 60 °C, in which the current density was controlled at 75 mA g⁻¹ within a cutoff voltage range of 2.0–4.0 V. After the cycling tests, the cells were decomposed in a dry room with an ultra-low humidity environment. Then, the electrolyte and electrodes were prudentially collected and respectively removed to the LC-MS (Nexera, Shimadzu Corp. and Esquire 3000 plus, Bruker) and DART-MS (DART-SVP, AMR Inc. and LCMS-8030, Shimadzu Corp.) equipments to carry out the electrolyte and electrode surface analysis. The details in the DART-MS measurement can be referred to our previous study.²⁵ In addition, four ion beam temperatures of 150, 250, 350 and 450 °C were used to detect the surface compounds of the electrodes.

Table 1 Various solvent combinations

Sample no.	Solvent	Additive
(a)	EC/DEC (1 : 1 v/v)	None
(b)	EC/DEC (1 : 1 v/v)	VC 1 wt%
(c)	PC/DEC (1 : 1 v/v)	None
(d)	PC/DEC (1 : 1 v/v)	VC 1 wt%

---

Results and discussion

Electrochemical characteristics of LiFePO₄/HC cells

Fig. 1 represents the cycle performance of the LiFePO₄/HC cells at 60 °C. Clearly, the cells with EC/DEC solvents, (a) and (b), show better cycle stability than PC/DEC, (c) and (d). The capacity retention at the 100th cycle for EC/DEC based solvent is over 68%; while, it is lower than 40% for PC/DEC. More importantly, VC substantially enhances the cycle performance for both the EC/DEC and PC/DEC cases, as shown in (b) and (d). It is also worth noting that the rate of capacity fading for the PC/DEC based solvent is particularly faster in the initial cycling, leading to the lower capacity retention. Accordingly, the electrochemical phenomena occurring in the initial cycles, i.e. the formation step should dominate the cell's cycle stability.

Fig. 1 Cycle performance at 60 °C for LiFePO₄/HC cells with various electrolytes. (a) EC/DEC (1 : 1 v : v); (b) EC/DEC (1 : 1 v : v) + 1 wt% VC; (c) PC/DEC (1 : 1 v : v); (d) PC/DEC (1 : 1 v : v) + 1 wt% VC.

Fig. 2 shows the 1st and 2nd charge/discharge curves for the cells of (a) EC/DEC and (b) PC/DEC solvents. For (a), a first charge capacity of ca. 160 mA h g⁻¹ can be achieved at 4 V; while, the first discharge capacity decreases to ca. 120 mA h g⁻¹, irrespective of the VC additive. Thus, it leads to a coulombic efficiency of 71%, which should be directly attributed to the intrinsic property of the hard carbon. In addition, a small peak is found at ca. 2.6 V, indicated by an arrow, for the case without VC; while, this peak shrinks and shifts toward lower charge voltage as the VC is added, reducing the charge polarization. The 2nd charge and discharge processes tend to balance with an enhanced coulombic efficiency, but their discharge capacities slightly decrease. The Fig. 2(b) shows a similar trend in the charge/discharge curves with (a), except for the peak during the initial charge step. It is indicated that the reaction behavior occurring in the EC/DEC solvent at the initial charge voltage differs from that of the PC/DEC solvent. Moreover, adding VC in the EC/DEC solvent can inhibit this initial reaction, on the other hand, VC has no obvious influence for the PC/DEC solvent.

Fig. 2 Initial charge/discharge curves for LiFePO₄/HC cells. (a) EC/DEC solvent; (b) PC/DEC solvent.

Analysis of electrolyte using LC-MS

Fig. 3 shows the LC-MS chromatograms of various electrolytes collected from the cells after the 2 cycle charge/discharge tests. Based on the identification results, the proposed chemical compounds are summarized in the bottom. Six compounds including P1–P3 (group P) and C1–C3 (group C) are found in the electrolyte composed of EC/DEC solvent, as shown in (a). Once VC is added (b), the peaks of P1–P3 and C2 disappear, and the peak intensity of C1 and C3 reduces to a certain level. On the other hand, for the PC/DEC solvents, (c) and (d), no chemical compounds can be identified. The compounds of group P should be attributed to the decomposition of LiPF₆; while, the compounds of group C with carbonate structure should form based on the decomposition of the EC/DEC solvent.^9,25 The decomposition reactions for these compounds are considered to correspond to the peak appearing in the 1st charge curve of Fig. 2(a). Also, the reduced decomposition products in Fig. 3(b) are related to the peak variation in intensity and charge voltage when VC is present. This means that VC inclines to inhibit the formation of both the decomposition products of group P and C. However, such decomposition reaction is hardly recognized by LC-MS analysis when the solvent is changed from EC/DEC to PC/DEC.

Fig. 3 LC-MS chromatograms of electrolytes after 2 cycle charge/discharge test.

Analysis of electrode surface using DART-MS

The cycled LiFePO₄ cathodes are analyzed by DART-MS and no chemical compound is found, except for the solvents, which is similar to the previously reported results.^9,24,25 For the HC anodes, several chemical compounds of group P and C in addition to the solvents are identified at various beam temperatures, as shown in Fig. 4. The chemical compounds formed on the anode surface for the EC/DEC solvent are very different from the PC/DEC. It is known from the case (a) and (b) that addition of VC induces the formation of P5 and P6, which are detected at the highest beam temperature of 450 °C. On the other hand for the cases (c) and (d), VC is found to assist the formation of other organophosphates P10 and P11, which are also found at the highest beam temperature. It is suggested that VC demonstrates different functions for the EC/DEC and PC/DEC solvents to form different types of thermally resistant organophosphates, as summarized in Fig. 5.

Fig. 4 Summary of DART-MS spectra of cycled HC anode surfaces under various beam temperatures.

Fig. 5 Schematic summary of the VC-induced film formation for different solvent bases.

Reaction mechanisms for the thermally resistant organophosphates

The organophosphate P6 was previously reported as a thermally resistant product,²⁵ which can reduce the decomposition of LiPF₆ and solvents on the anode surface, thus enhancing the cycle stability. This is also in consistent with the less decomposition products in the electrolyte, as observed in Fig. 3(b). Compared with EC/DEC, the PC/DEC solvent tends to decompose to form not only the organophosphates but also the carbonate oligomer C4 on the HC surface, corresponding to the relatively high rate of capacity fading during cycling (Fig. 1). However, this rapid capacity fading can still be reduced to a certain extent by adding VC, which may be attributed to the formation of the thermally resistant organophosphates P10 and P11. Fig. 6 shows the proposed reaction mechanisms for the formation of the thermally resistant organophosphates P6, P10 and P11. The details of the reaction mechanism for P6 can be referred to the previous literature.²⁵ For the other two mechanisms occurring in the PC/DEC solvent, either oxidation^30,31 or reduction³² of PC plays an important role in forming the organophosphates in addition to the decomposition of LiPF₆. This is very different from that in the EC/DEC solvent, where only the reduction of EC is taken into consideration,²⁵ thus resulting in the formation of different types of organophosphates.

Fig. 6 Proposed reaction mechanisms for the thermally resistant organophosphates.

Conclusions

The film formation induced by VC additive in the LiFePO₄/HC cells at 60 °C is investigated via LC-MS together with DART-MS. The cells with EC/DEC solvent show better cycle performance than PC/DEC at 60 °C and the addition of VC substantially improves their cycle performances. According to the initial charge/discharge behaviors, VC is found to suppress and shift the peak appearing in the 1st charge curve of the cell with EC/DEC solvent, while no peak is observed and the charge/discharge curves are almost the same for the PC/DEC solvent, irrespective of VC. Based on the LC-MS and DART-MS analysis, VC is also found to induce the formation of different types of thermally resistant organophosphates on the HC surface for the EC/DEC and PC/DEC solvents. These thermally stable compounds could prevent further degradation of electrolyte at elevated temperature, thus enhancing the cell's cycle performance.

Note added after first publication

This article replaces the version published on 10th August 2016, in which an incorrect version of Figure 4 was presented through editorial error.

Acknowledgements

The authors would like to appreciate Ms Miki Okano of National Institute of Advanced Industrial Science and Technology (AIST) for the technical support on the experiment.

Notes and references

1. T. Tanaka, K. Ohta and N. Arai, J. Power Sources, 2001, 97–98, 2–6.
2. R. Hamlen, G. Au, M. Brundage, M. Hendrickson, E. Plichta, S. Slane and J. Barbarello, J. Power Sources, 2001, 97–98, 22–24.
3. R. A. Marsh, S. Vukson, S. Surampudi, B. V. Ratnakumar, M. C. Smart, M. Manzo and P. J. Dalton, J. Power Sources, 2001, 97–98, 25–27.
4. A. Chu and P. Braatz, J. Power Sources, 2002, 112, 236–246.
5. T. Takamura, Solid State Ionics, 2002, 152–153, 19–34.
6. H. Xia, K. R. Ragavendran, J. Xie and L. Lu, J. Power Sources, 2012, 212, 28–34.
7. X. Tang, S. S. Jan, Y. Qian, H. Xia, J. Ni, S. V. Savilov and S. M. Aldoshin, Sci. Rep., 2015, 5, 11958.
8. H. Xia, Q. Xia, B. Lin, J. Zhu, J. K. Seo and Y. S. Meng, Nano Energy, 2016, 22, 475–482.
9. S. Takeda, W. Morimura, Y. H. Liu, T. Sakai and Y. Saito, Rapid Commun. Mass Spectrom., 2016, 30, 1754–1762.
10. M. Herstedt, H. Rensmo, H. Siegbahn and K. Edström, Electrochim. Acta, 2004, 49, 2351–2359.
11. J. Vetter, M. Holzapfel, A. Wuersig, W. Scheifele, J. Ufheil and P. Novák, J. Power Sources, 2006, 159, 277–281.
12. L. El Ouatani, R. Dedryvère, C. Siret, P. Biensan, S. Reynaud, P. Iratçabal and D. Gonbeau, J. Electrochem. Soc., 2009, 156, A103–A113.
13. I. B. Stojković, N. D. Cvjetićanin and S. V. Mentus, Electrochem. Commun., 2010, 12, 371–373.
14. D. Xiong, J. C. Burns, A. J. Smith, N. Sinha and J. R. Dahn, J. Electrochem. Soc., 2011, 158, A1431–A1435.
15. J. C. Burns, N. N. Sinha, D. J. Coyle, G. Jain, C. M. VanElzen, W. M. Lamanna, A. Xiao, E. Scott, J. P. Gardner and J. R. Dahn, J. Electrochem. Soc., 2012, 159, A85–A90.
16. D. Takamatsu, Y. Orikasa, S. Mori, T. Nakatsutsumi, K. Yamamoto, Y. Koyama, T. Minato, T. Hirano, H. Tanida, H. Arai, Y. Uchimoto and Z. Ogumi, J. Phys. Chem. C, 2015, 119, 9791–9797.
17. C. Jehoulet, P. Biensan, J. M. Bodet, M. Broussely, C. Moteau and C. Tessier-Lescourret, Proc. Electrochem. Soc., 97-18, The Electrochem. Soc. Inc., Pennington, NJ, 1997, pp. 974–985.
18. S.-K. Jeong, M. Inaba, R. Mogi, Y. Iriyama, T. Abe and Z. Ogumi, Langmuir, 2001, 17, 8281–8286.
19. X. R. Zhang, R. Kostecki, T. J. Richardson, J. K. Pugh and P. N. Ross, J. Electrochem. Soc., 2001, 148, A1341–A1345.
20. R. Oesten, U. Heider and M. Schmidt, Solid State Ionics, 2002, 148, 391–397.
21. D. Aurbach, K. Gamolsky, B. Markovsky, Y. Gofer, M. Schmidt and U. Heider, Electrochim. Acta, 2002, 47, 1423–1439.
22. K. Xu, Chem. Rev., 2004, 104, 4303–4417.
23. L. Chen, K. Wang, X. Xie and J. Xie, J. Power Sources, 2007, 174, 538–543.
24. L. El Ouatani, R. Dedryvère, C. Siret, P. Biensan and D. Gonbeau, J. Electrochem. Soc., 2009, 156, A468–A477.
25. Y. H. Liu, S. Takeda, I. Kaneko, H. Yoshitake, M. Yanagida, Y. Saito and T. Sakai, Electrochem. Commun., 2015, 61, 70–73.
26. T. Zheng, J. S. Xue and J. R. Dahn, Chem. Mater., 1996, 8, 389–393.
27. R. Yazami and M. Deschamps, J. Power Sources, 1995, 54, 411–415.
28. T. Zheng, W. Xing and J. R. Dahn, Carbon, 1996, 34, 1501–1507.
29. Q. Zhang, H. Sun, X. Wang, Z. Zhu, W. Liang, A. Li, S. Wen and W. Deng, Energy Technol., 2013, 1, 721–725.
30. L. Xing, C. Wang, W. Li, M. Xu, X. Meng and S. Zhao, J. Phys. Chem. B, 2009, 113, 5181–5187.
31. E. G. Leggesse, R. T. Lin, T. F. Teng, C. L. Chen and J. C. Jiang, J. Phys. Chem. A, 2013, 117, 7959–7969.
32. X. Zhang, R. Kostecki, T. J. Richardson, J. K. Pugh and P. N. Ross Jr, J. Electrochem. Soc., 2001, 148, A1341–A1345.

---