The amorphisation of the lithiated crystalline silicon after the extraction of lithium in LiPF6/carbonates mixture based electrolyte has been demonstrated by the work of Obrovac and Christiansen [1] ten years ago.Within this work, we focus on the structural changes of the crystalline nanosilicon (nSi, 30-50nm) during electrochemical cycling in ionic liquid (IL) based electrolytes. By means of Raman Spectroscopy and X-Rays Diffraction we investigate the nSi electrodes after 1 cycle of lithium insertion/extraction in 0.8 M LiTFSI in 1-Methyl-1-propylpyrrolidiniumbis (trifluoromethylsulfonyl) imide (P13TFSI) and 1-Methyl-1-propylpiperidiniumbis (trifluoromethylsulfonyl) imide (PP13TFSI). For comparison, nSi electrode is also cycled in conventional LP30 electrolyte. The electrodes consist of 85 wt-% of nSi, 5 wt-% carbon black and 10 wt-% of CMC, the electrode is loading about 0.2 mg*cm-2, insertion/extraction current is 180 mA*g-1.During the first lithium insertion/extraction similar electrochemical behavior is observed in all electrolytes (Fig.1). Nevertheless, during extended cycling the capacity fade is much faster in ionic liquids with respect to LP30. The Raman spectra (Fig. 2) taken after lithium extraction in LP30 show clearly the broad band of amorphous silicon between 430 – 500 cm-1. In contrary no amorphous silicon is found after lithium extraction from nSi electrode in ionic liquids. Instead, the band located at ~ 506 cm-1 is registered in reproducible manner. The sample heating effect leading to the silicon phase transformation and peak of the shift can be excluded as low power of the laser was applied (less than 3 mW) and no significant difference was registered in three subsequent measurements. This shift has already been observed for lithiated silicon, however it has not been commented [2]. On the other hand, very detailed Raman spectroscopy investigation of the phase evolution of crystalline cubic silicon Si-I evolution under stress by means of Vickers indentation has been performed [3] [4]. In the center of the indent, amorphisation of the silicon leading to the broad Raman band at ~480 cm-1 was observed, while a pileup at the border of the indent was assigned to hexagonal silicon Si-IV, appearing as a band at 500-510 cm-1in Raman spectra. Direct phase transition of Si -I to Si-IV requires a high level of a shear deformation. Since the same shift of the Raman peak is observed in our electrochemical experiments of lithiation/delithiation of crystalline cubic nSi, it could be intuitively assumed that hexagonal silicon Si-IV is formed in ionic after cycling in ionic liquids. The XRD measurements confirm Raman Spectroscopy findings. The investigation, if the observed structural differences are the reason of the faster capacity fading of silicon electrode in ILs, is in progress.Is it possible that silicon is exposed to high shear deformation when cycled in ILs? Is it related to the insertion/reaction of TFSI anion which takes place at 0.8 – 1V? Within this work we try at least partially answer these questions. To our best knowledge the basic characteristic of silicon behavior in ILs has never been addressed in electrochemical setups in such context. Therefore, taking into account the interest of the scientific community towards application of silicon anodes as well as ionic liquids electrolytes, we consider this basic study of high importance.Fig. 1 Lithium insertion - extraction curves into nSi electrode recorded in LP30, P13TFSI and PP13TFSI, 180 mAg-1 Fig. 2 Raman Spectra of the nSi electrodes, pristine and after one cycle (Fig.1).[1] M.N. Obrovac, L. Christensen, Electrochemical and Solid-State Letters, 7 (2004) A93-A96.[2] H. Li, X. Huang, L. Chen, G. Zhou, Z. Zhang, D. Yu, Y.J. Mo, N. Pei, Solid State Ionics, 135 (2000) 181-191.[3] A. Kailer, Y.G. Gogotsi, K.G. Nickel, Journal of Applied Physics, 81 (1997) 3057-3063.[4] T. Wermelinger, R. Spolenak, Journal of Raman Spectroscopy, 40 (2009) 679-686.
Read full abstract