Elucidation of Reaction Behavior of Phosphorus-Doped Si Negative Electrode for Lithium-Ion Batteries

Abstract

Silicon (Si) is a promising active material as a negative electrode in next-generation lithium-ion batteries (LIBs) due to its high theoretical capacity of 3580 mA h g− 1 (Li15Si4). It has been reported that crystalline Si (c-Si) forms an amorphous Li-Si alloy phase (a-Li x Si) in the first charge process and that the resulting a-Li x Si subsequently alloys with Li to form a c-Li15Si4 phase.1 The volumetric change ratio per Si atom from c-Si to c-Li15Si4 corresponds to 380%, which generates high stresses and large strains in the active materials. The strains that accumulate under repeated charge−discharge cycling cause disintegration of the Si negative electrode, leading to a rapid decrease in capacity. Si also has disadvantages of a low electrical conductivity and a low diffusion coefficient of Li+ within it. To address these issues, doping of Si with impurities has been attempted to increase the electrical conductivity of Si. However, the reaction behavior of the doped Si has not yet been clarified. In this study, the relationship between the electrochemical performance of phosphorus (P)-doped Si negative electrode and the reaction behavior was investigated by X-ray diffraction (XRD), Raman spectroscopy, field emission scanning electron microscopy (FE-SEM), and so on. P-doped Si (P concentration; 0, 50, 124, 633, and 1000 ppm) powders were supplied by Elkem (Sigrain e-Si). Electrochemical measurements were carried out with a laboratory-made three-electrode cell. The working electrode was fabricated by gas-deposition method, which is without any binder or conductive agent. Both the counter and reference electrodes consisted of Li metal. The electrolyte solution used was 1 M lithium bis(trifluoromethanesulfonyl)amide (LiTFSA) in propylene carbonate (PC). XRD and Raman spectroscopy were used to investigate the phase transition and change in the surface crystallinity of P-doped Si electrodes before and after a charge-discharge cycle. The surface morphology and cross-sectional surface of the P-doped Si electrodes were observed by FE-SEM. A focused ion-beam was used to fabricate the cross-sectional surface. Li insertion energy (E insertion) into Si was also calculated based on computational chemistry. The electric resistivity and the first charge-discharge capacity of a Si electrode decreased under P-doping; an increase in electrical conductivity does not always correspond to an increase in the initial charge-discharge capacity. In addition, a very low P concentration of 124 ppm has a significant influence on the electrochemical properties of a Si negative electrode; P-doping of Si suppressed the rapid fading of the discharge capacity. After the charge-discharge cycle, the undoped Si electrode was full of cracks and the active material of Si was deteriorated. On the other hand, the surface morphology of P-doped Si was almost the same as that before cycle, even though a slight crack was observed. Therefore, P-doping suppressed the change in the surface morphology of a Si negative electrode after charge-discharge cycles. It is well-known that c-Si has a triply degenerated Raman active F2g-mode at around 520 cm-1, and that this peak shifts toward a lower wavenumber at around 490 cm-1. The peak assigned to F2g-mode of P-doped Si electrode was shifted to lower wavenumber, and the degree of the peak shift increased with a decrease in the P concentration; P-doping suppressed the change from c-Si to amorphous Si (a-Si). While the peaks assigned to c-Si phase disappeared in XRD patterns of 0 and 50 ppm P-doped Si electrodes after the first lithiation at 0.005 V vs. Li+/Li in LiTFSA/PC, these peaks remained in XRD patterns of 124, 633, and 1000 ppm P-doped Si electrodes. In addition, the intensity of peaks which attribute to c-Li15Si4 phase decreased with an increase in the P concentration. Hence, P-doping controlled the formation of a c-Li15Si4 phase during charge process. This suppression of the phase transition led to a decrease in the large change in volume of Si and prevented the Si negative electrode from disintegrating. E insertion increased with an increase in the P concentration; Li insertion into P-doped Si is energetically unfavorable, which indicates that the crystal lattice of Si shrinks as a result of the replacement of some Si atoms with smaller P atoms. Hence, it is more difficult to insert into P-doped Si. These effects of P-doping help to improve the otherwise poor cycle performance of a Si electrode.2 References1) J. R. Dahn, et al., J. Alloy Compd., 2010, 496, 25-36.2) Y. Domi, H. Usui, H. Sakaguchi, et al., ACS Appl. Mater. Interfaces, 2016, 8, 7125-7132. AcknowledgmentsThis work was partially supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI Grant numbers 24350094 and 15K21166.