Introduction

The limited capacity of graphite which is the most widely used anode material for lithium ion batteries has provoked numerous studied seeking alternative materials. Much attention has been given to Si-based anode materials because silicon has the high specific capacity of 4200 mAh g⁻¹. However, silicon undergoes a change in volume during charging and discharging, and this results in mechanical instability and poor cycle life performance [1,2]. The mechanical stresses induced by the large volumetric change should be buffered in order to obtain better cyclability with silicon-based anodes. Various carbonaceous materials such as graphite [3-7], carbon nanofiber [8-11] and amorphous carbon [12-15] have been suggested as an electrochemically active component which constitutes a composite with Si for buffering the volumetric change. Several metallic compounds with high mechanical strength and electronic conductivity have been also proposed to suppress Si volume expansion [16-20].

Porous Si is also considered as one of the candidates to improve the cycling stability. Zhihao et al. reported microporous silicon replicas with an interconnected network of silicon nanocrystals and 3-dimensional frustule morphology by simple magnesium reduction method [21]. Chen et al. synthesized Si/MgO composites with reasonable cycling stability by magnesium reduction method [22]. Recently, Ying et al. investigated the electrochemical performances of nano-porous Si prepared by an aluminum redox reaction [23]. The nano-porous Si/graphite/C composite exhibited a reversible capacity of about 700 mAhg⁻¹ with no capacity loss up to 120^th cycle.

In this work, we investigated the electrochemical performances of the porous Si + C composite prepared by mechanical milling of SiO with magnesium powder followed by pyrolysis of pitch. Another composite containing nonporous Si was prepared by same procedure and the properties were compared with those of the porous Si + C composite.

Material and Methods

2.1. Synthesis of materials

Si/MgO composite powder was synthesized by mechani- cal alloying of SiO (Osaka titanium Tech., Japan) and magnesium powder (Samchun pure chemical Co., Korea) with a SPEX mill. The molar ratio of SiO:Mg was 1:1 and milling was conducted at 400 rpm for 20 h in argon atmosphere. The weight ratio of zirconia ball (3 mm) to mixture was 10:1. Magnesium was used as a reducing agent. The mixture was loaded in the hardened steel bowl and sealed in an argon filled glove box. The powder obtained after the milling was put into hydrochloric acid solution while stirring. The etching process lasted for 8 h followed by washing, filtering and drying under vacuum at 80℃ for 24 h. By this process, MgO in the composite was dissolved and nano-porous Si (denoted as PSi hereafter) was formed.

The obtained PSi was added in a pitch solution in tetrahydrofuran (THF), and dispersed ultrasonically. The pitch solution containing PSi was dried at 80℃ in a convection oven for 12 h and finally calcined at 900℃ under Ar flow for pyrolysis of the pitch. Nonporous microsized Si (B&C, Japan) (denoted as MSi hereafter) underwent the same procedure to prepare MSi + C composite for reference.

2.2. Analysis

The crystal structure of the obtained powder was examined by powder X-ray diffraction (XRD, Rigaku D/Max- 2500/PC, Japan) using a Cu Kα radiation source. The particle size was analyzed using a particle size analyzer (PSA, Partica LA-950V2, Horiba, Japan) and the particle morphology was observed by field emission scanning microscopy (FE-SEM, S-800, Hitachi, Japan).

2.3. Electrochemical performance test

A coin-cell (CR2016) was used to determine the electrochemical performance of the obtained samples. A negative electrode was made by coating a slurry of the active material, Ketjen black and a polyvinylidene fluoride (PVdF) binder (Kurea, Japan) at a weight ratio of 80:10:10 onto a copper foil current collector. The electrolyte used was a 1.0 M LiPF₆ solution in ethylene carbonate/ethyl-methyl carbonate (EC/EMC) (1/1 vol%) (Cheil Ind., Korea). The galvanostatic charge and discharge cycle tests of the cell were carried out at room temperature between 0.01 V and 1.5 V. The cell was charged (lithiation) with constant current-constant voltage (CC-CV) mode at 0.1C with cut-off current of 0.01C and discharged (de-lithiation) with CC mode at 0.1C for formation cycle. This cycling procedure was repeated 50 times to test cycling performance of the sample. Cycling volammetry and impedence spectroscopy (EIS) of the cells was measured using a potentialstat (VSP, Biologic, France).

Results and Discussion

Fig. 1 presents the XRD pattern of the Si/MgO mixture obtained after mechanical alloying and PSi obtained after the etching process. The strong peaks in Fig. 1(a) correspond to Si and MgO, and no peaks of metallic magnesium are identified. In addition, some weak peaks are detected, which are assignable to Mg₂SiO₄. These materials should be the by-products of reaction (1).

Fig. 1.

X-ray diffraction (XRD) pattern of (a) Si/MgO and (b) porous Si (PSi) obtained after the etching process.

(1)

SiO + Mg → Si + MgO

In Fig. 1(b), diffraction peaks for MgO and Mg₂SiO₄ are not observed which indicates these compounds were almost removed in the etching process. Considering the intensity and sharpness of the peak in this figure, Si particles with high crystallinity was obtained in this process.

The SEM images of PSi is given in Fig. 2. Sub-micrometer sized particles are aggregated to form secondary particles. Looking into the surface of the particles with higher magnification, nanometer sized particles are observed to be aggregated and the mesopores are supposed to exist among the nanoparticles.

Fig. 2.

Scanning electron microscope (SEM) images of PSi.

Specific surface area (SSA) measured by Brunauer- Emmett-Teller (BET) method are listed in Table 1. The SSA for PSi is 64.52 m²g⁻¹, which is much larger than 4.28 m²g⁻¹ for the Si/MgO composite. The BJH pore size distribution for PSi is shown in Fig. 3. Most pores have diameter ranging from 9 to 15 nm showing 9.31 nm of d_p value. These pores are believed to be formed from removal of MgO nanoparticles embedded in the Si/MgO composites.

Fig. 3.

BJH pore size distribution of PSi.

Table 1.

The specific surface area of the Si/MgO composite before and after the etching

Sample	Surface area (m2g–1)	Pore size (nm)
Before etching	4.28	21.21
After etching	64.52	9.31

Assuming no reaction between the pitch and Si, the silicon content in the PSi/C composite is calculated to be 18.2 wt%.

Fig. 4 presents the charge and discharge profile of composites (PSi/C and MSi/C) at the first cycle. PSi/C composite shows the first charging capacity of 1012 mAh g⁻¹ and discharging capacity of 607 mAh g⁻¹ (Coulombic efficiency = 60.0%) while MSi/C composite exhibits 1300 mAh g⁻¹ and 879 mAh g⁻¹ (Coulombic efficiency = 67.6%) for the first charge and discharge respectively. The first charge and discharge capacity of PSi/C composite is less than that of MSi/C, which may be due to hidden impurities – possibly amorphous oxidized product of SiO_x generated from the milling process – unidentified from XRD analysis. PSi/C composite shows lower Coulombic efficiency than MSi/C composite. In general, electrode material with large surface area has low Coulombic efficiency because of active side reaction consuming lithium ion. It should be noted that the charge capacity represented at the interval of CC-mode is larger for PSi/C than MSi/C composite, which indicates the resistance in lithiation process is lower for PSi/C than MSi/C composite. The shorter lithium diffusion length of PSi/C composite together with enhanced electronic conductivity due to carbon in the pores of silicon can be the reasons for the lower resistance in lithiation. Cycling stability of the composites is shown in Fig. 5. The reversible capacities of PSi/C and MSi/C composite retained after 50 cycles are 378 mAh g⁻¹ (59.6% capacity retention) and 255 mAh g⁻¹ (28.0% capacity retention). These cycling stability curves distinctly show that the capacity fading of MSi/C composite electrode is much more mitigated comparing with PSi/C composite electrode. The reason may be that the pore on surface of PSi can act as a buffer to accommodate a certain amount of volume change. TheCoulombic efficiency of the electrode becomes close to 100% from the 2^nd cycle for both the composites.

Fig. 4.

Charge-discharge profile of the 1^st cycle of the PSi/C and MSi/C composite.

Fig. 5.

Cycling stability of the PSi/C and MSi/C composite. (current density: 0.1C charge/0.1C discharge)

It should be noted that PSi/C electrode begins to show capacity fading after 20^th cycle while MSi/C electrode shows monotonous fading from the beginning Microcracks seems to be developed until 20^th cycle and macrocracks which degrade the cycleability are developed after that point in case of PSi/C electrode. However, macrocracks appear to be developed from the initial in case of MSi/C electrode because there is no buffer to absorb the volumetric change during cycling.

The microstructural changes before and after the cycling tests are seen in the SEM images of PSi/C and MSi/C (Fig. 6). We can see that the composite PSi/C electrode keeps the homogeneous surface morphology and shows less cracks than MSi/C electrode even after 50 cycles. The structure of PSi/C composite is more endurable than MSi/C composite against volumetric change of Si during charge and discharge and this structural stability might give less cracks. The cracks at the surface of MSi/C electrode are expected to break electronic conductive path and cause severe capacity fading.

Fig. 6.

The morphological change of the composite electrodes: (a) PSi and (b) MSi after 50 cycles.

The cyclic voltammograms in Fig. 7 show electrochemical behavior of the composites during charging and discharging in the potential range from 0.0 to 3.0V. No peaks around 0.6 to 1.5 V which represent electrolyte decomposition did not appear for both the composite. In the case of PSi/C, a cathodic current is observed at about 0.3 V at the first and second scanning, which corresponds to electrochemically driven amorphization, while no peaks are found with MSi/C composite at this region. The 50^th scanning shows no obvious cathodic or anodic current peaks at certain voltage but weak current all over the range for both the composite.

Fig. 7.

Cyclic voltammogram of (a) PSi/C and (b) MSi/C composite.

Electrochemical impedance spectroscopy (EIS) data measured at frequency from 10 mHz to 100 kHz are shown in Fig. 8. The size of the semicircles is similar for the two composites at initial stage (after 1^st cycle) but MSi/C composite shows larger increase in the size after 51^st cycle. It means that increase in polarization resistance of MSi/C electrode is larger than that of PSi/C composite, which may be attributed to less degradation of the PSi/C electrode due to cycling stability. Resistance in lithium diffusion can be estimated from the EIS curve of the cycled electrodes at low frequency range. PSi/C electrode exhibits lower resistance than MSi/C electrode, which is attributed to shorter lithium diffusion length of the porous structure of PSi/C composite and crack formation in MSi/C electrode during the cycling.

Fig. 8.

Electrochemical impedance spectra (EIS) of (a) PSi/C and (b) MSi/C composite.

Conclusions

The effects of porous structure on the cycling stability of Si as an anode material were investigated by preparing a composite of porous Si + carbon. The nano-porous Si was prepared by a simple magnesium reduction process and the etching process gave mesoporous Si with mean pore diameter of 9.31 nm. The BET surface area of porous Si (64.52 m²g⁻¹) was extremely higher than that of before etching Si/MgO (4.28 m²g⁻¹). The electrode prepared with the porous Si (PSi/C composite) exhibited improved cycling stability, compared with micro-sized Si (MSi/C). The PSi/C composite retained 59.6% of the first cycle after 50 cycles while MSi/C showed only 28.0% of the capacity retention. It is believed that the pores of Si formed after removal of MgO and Mg₂SiO₄ can accommodate the large volumetric change of Si during charge/ discharge process.

Acknowledgements

Acknowledgements

This study was supported by a grant from Chungcheong leading industry promotion project of the Korean Ministry of Knowledge Economy.

REFERENCES

• 1. T. D Besenhard, J Yang and M Winter: J. Power Sources., (1997) 68 87.Article
• 2. C. J Wen and R. A Huggins: J. Solid State Chem., (1976) 37 271.Article
• 3. A. M Wilson and J. R Dahn: J. Electrochem. Soc., (1995) 142 326.ArticlePDF
• 4. H Li, X Huang, L Chen, Z Wu and Y Liang: Electrochem. Solid-State Lett., (1999) 2 547.Article
• 5. I.-S Kim, P. N Kumta and G. E Blomgren: Electrochem. Solid-State Lett., (2000) 3 493.
• 6. I.-S Kim, G. E Blomgren and P. N Kumta: J. Power Sources., (2004) 130 275.Article
• 7. M Yoshio, H Wang, K Fukuda, T Umeno, N Dimov and Z Ogumi: J. Electrochem. Soc., (2002) 149 A1598.Article
• 8. N Dimov, S Kugino and M Yoshio: Electrochim. Acta., (2003) 48 1579.Article
• 9. J Yang, B. F Wang, K Wang, Y Liu, J. Y Xie and Z. S Wen: Electrochem. Solid-State Lett., (2003) 6 A154.Article
• 10. H Dong, X. P Ai and H. X Yang: Electrochem. Commun., (2003) 5 952.Article
• 11. H. Y Lee and S. M Lee: J. Power Sources., (2002) 112 649.Article
• 12. Y Liu, K Hanai, T Matsumura, N Imanishi, A Hirano and Y Takeda: Electrochem. Solid-State Lett., (2004) 7 A492.Article
• 13. M. S Park, Y. J Lee, S Rajendran, M. S Song, H. S Kim and J. Y Lee: Electrochim. Acta., (2005) 50 5561.Article
• 14. H. Y Lee, Y. L Kim, M. K Hong and S. M Lee: J. Power Sources., (2005) 141 159.Article
• 15. J Yang, Y Takeda, N Imanishi, C Capiglia, J. Y Xie and O Yamamoto: Solid State Ionics., (2002) 152/153 125.
• 16. X Yang, J McBreen, W Yoon, M Yoshio, H Wang, K Fukuda and T Umeno: Electrochem. Commun., (2002) 4 893.Article
• 17. N Dimov, S Kugino and M Yoshio: J. Power Sources., (2004) 136 108.Article
• 18. G. X Wang, J Yao and H. K Liu: Electrochem. Solid- State Lett., (2004) 7 A250.Article
• 19. T Kim, S Park and S. M Oh: Electrochem. Commun., (2006) 8 1461.Article
• 20. J. H Kim, H Kim and H. J Sohn: Electrochem. Commun., (2005) 7 557.Article
• 21. B. C Kim, H Uono, T Sato, T Fuse, T Ishihara and M Senna: Solid State Ionics., (2004) 172 33.Article
• 22. T Morita and N Takami: J. Electrochem. Soc., (2006) 153 A425.Article
• 23. B. C Kim, H Uono, T Satou, T Fuse, T Ishihara, M Ue and M Senna: J. Electrochem. Soc., (2005) 152 A523.Article

Figure & Data

References

Citations

Citations to this article as recorded by