Preparation of Carbon-Silicon Composites by Potassiothermic Reduction of Silicon Tetrachloride and Their Electrochemical Performance
Article information
Abstract
The traditional graphite electrode materials can't meet the needs of the Li-ion batteries with high energy density because of its low specific capacity. It is of great practical significance to modify the traditional graphite electrode materials and improve their specific capacity. In this paper, a simple potassiothermic reduction of silicon tetrachloride method is used to recombine silicon onto the conventional graphite electrode materials, which greatly improves the specific capacity and the cycling performance of the electrode materials. When the content of silicon in the carbon-silicon (C/Si) composites is 8.5%, the theoretical specific capacity of the obtained C/Si composites is 644.6 mAh g–1, and after 1000 cycles of charge and discharge, the actual specific capacity is 570 mAh g–1, which is about 88.4% of the theoretical specific capacity; when the content of silicon in the C/Si composites is 19.0%, the theoretical specific capacity of the obtained C/Si composites is 981.33 mAh g–1, and after 1000 cycles of charge and discharge, the actual specific capacity can still reach 820 mAh g–1, which is about 83.6% of the theoretical specific capacity; when the content of silicon in the C/Si composites is 28.7%, the theoretical specific capacity of the obtained C/Si composites is 1292.4 mAh g–1, and after 1000 cycles of charge and discharge, the actual specific capacity is 710 mAh g–1, which is about 54.9% of the theoretical specific capacity; therefore, the optimal content of silicon in C/Si composites is about 19.0%. Through incorporating silicon into traditional graphite electrode materials via potassiothermic reduction of silicon tetrachloride, the specific capacity of the negative electrode is improved significantly, which is a better way to upgrade the products of the traditional graphite negative electrode materials manufacturers.
INTRODUCTION
The traditional anode material (graphite) is confronting a great challenge due to its fairly low specific capacity (372 mAh g–1) and potential safety concerns in connection with the growth of lithium dendrites [1,2]. Silicon (Si) has been considered as a promising anode material because of its abundant reserves in nature, low lithium ion (Li+) intercalation/de-intercalation potential (below 0.5 V vs. Li/Li+) and high theoretical capacity of 4200 mAh g–1 [3–5]. Despite these outstanding merits, Si anode has not yet been commercially applied owing to the following two technical challenges. First, the electronic conductivity of the Si is poor, which could not bear the charge/discharge at heavy current density [6]. Second, pure Si suffers from huge volume expansion/shrinkage (over 300%) during the Li insertion/extraction process, which leads to the dramatic crack of the Si particle and pulverization from the current collector [7]. Therefore, there is an urgent need to suppress the volume expansion and enhance the conductivity of the Si electrode, so as to improve its cycle life, and rate capability [8]. Hence, tremendous strategies have been devoted to alleviate the volume changes and enhancing the electrochemical performance of Si anodes. Nanostructuring strategies for Si, such as nanoparticles [9], nanowires [10], nanotubes [11], nanosheets [12], and porous nanostructures [13,14] have been verified to be an effective way to optimize the cyclic performance. Carbon-silicon compositing has been considered as a simple and highly efficient way to boost the Li-storage performance of Si, due to the following ameliorating effects [15–17]: (i) carbon can be functionalized as a buffer to tolerate severe volume change of Si after electrochemical Li alloying reaction; (ii) ionic and electronic conducting ability of the Si anode can be greatly improved, thus promoting the electrochemical reaction kinetics of Si with Li. The simplest way to prepare silicon-carbon composite anodes is to mechanically mix carbon materials and nano-silicon. Due to the size difference between the two, nano-silicon can fill between graphite particles and utilize pores to buffer volume changes, while the matrix of carbon materials can mechanically buffer volume changes. As early as 1998, Wang et al. [18] used a ball milling method to mechanically mix nano-silicon particles with graphite to prepare silicon-carbon anode materials. The initial specific capacity could reach 1039 mAh g–1, but only 794 mAh g–1 was maintained after 20 cycles. This is because nano-silicon is prone to agglomeration and there is no interaction force between it and graphite. During the lithium intercalation and deintercalation process, the significant volume change difference between nano-silicon and carbon materials still leads to a rapid loss of electrical contact, resulting in capacity fading. Li et al. [19] used a ball milling method to ground sawdust from the photovoltaic industry to obtain nano-silicon particles. Then, after removing metal impurities with hydrochloric acid and organic impurities with heat treatment, they mixed with phenolic resin to obtain the precursor. Finally, they successfully prepared nano-silicon-carbon composites with a carbon-coated shell-core structure by hightemperature pyrolysis of phenolic resin. The core-shell structure enhanced the conductivity of the silicon-carbon anode, promotes the movement of ions/electrons, and reduced the volume fluctuation of silicon during the lithiation/delithiation process. Compared with pure silicon anodes, the Li+ transport efficiency of nano-silicon-carbon was significantly improved, demonstrating better cycling and rate performance. At a current density of 0.5 A g–1, after 150 cycles, the nano-silicon-carbon anode has a specific capacity 440 mAh g–1 higher than that of the pure silicon anode, demonstrating more excellent cycling performance. Kim et al. [20] used propane, a gas-phase carbon source, to pyrolyze-nucleate-grow and form films on the surface of nano-silicon particles at 700°C to obtain shell-core-structured silicon-carbon materials. After preparing the electrode, they fabricated holes inside the electrode sheet by laser ablation to construct a three-dimensional porous siliconcarbon composite electrode. After 25 cycles the nano-silicon-carbon anode maintained a specific capacity of more than 1500 mAh g–1, while the specific capacity of the pure silicon anode tended to zero, demonstrating the significant role of the carbon layer in enhancing the cycling performance of silicon anode. Xu et al. [21] used CMC as the dispersant to mix nano-silicon particles with a suspension of polyethylene pyroloranone and glucose, and added flake graphite. After ball grinding and dispersion, they spray-granulated into microsphere particles. After high-temperature pyrolysis and carbonization at 900°C, they used the gas-phase carbon deposition of acetylene to coat the carbon layer, and finally prepared a silicon-carbon anode material with a watermelon-like structure. In this structure, the nano-silicon core helped the anode achieve a specific capacity of 620 mAh g−1, demonstrating excellent capacity performance. The carbon framework and coating effectively prevent the erosion of the electrode solution. Moreover, the nano-sized silicon particles could be uniformly assembled into silicon-carbon microspheres instead of aggregates, further reducing the probability of electrolyte side reactions. As a result, a Coulombic efficiency of 89.2% was achieved during the first charge and discharge cycle.
In this work, a new carbon-silicon compositing method is proposed. The commercial graphite electrode material is used as the carbon source, and the carbon-silicon compositing is realized by the method of potassiothermic reduction of silicon tetrachloride, and good results are obtained. The silicon tetrachloride is reduced by metal potassium to form a thin silicon layer firmly attached to the commercial graphite electrode materials, and the only by-product potassium chloride is directly eliminated by water scrubbing. Therefore, the finally obtained C/Si has no other impurities. The obtained C/Si composites have a significant increase in specific capacity as compared with the raw materials used (commercial graphite anode materials). It is a feasible method for upgrading the products of traditional graphite anode material manufacturers.
EXPERIMENTAL
Materials and Equipment
Commercial graphite anode materials were purchased from BTR New Energy Material Ltd. Commercial LiFePO4 cathode materials were purchased from Hunan Yuneng New Energy Battery Material Co., Ltd. The high pressure reactor used for potassiothermic reduction of silicon tetrachloride is a self-made device as shown in Fig. 1. The inner diameter of the reactor is 16mm, and the working volume of the device is 25 mL.
The photographs of a carbon-silicon composite anode material synthesis device: (a) high pressure reactor disassembly diagram, (b) assembled high pressure reactor.
Preparation of C/Si Composites
Graphite, silicon tetrachloride and metal potassium were weighed according to the theoretical content of silicon in the carbon-silicon composite of 10%, 20% and 30%, and the actual weight of graphite, metal potassium and silicon tetrachloride weighed is shown in Table 1. Then they were placed in a self-made reactor and reacted at 300°C for 4 hours. After the reaction was completed, the as-prepared products were washed with a large amount of distilled water, then were dried in a 120°C incubator for 12 hours to obtain the C/Si composites. The main chemical reaction is as follows:
Weight of materials of graphite, silicon tetrachloride and metal potassium
Analysis of Chemical Compositions of C/Si Composites
The silicon content of the as-prepared C/Si composites was measured by chemical analysis. The specific steps were: weighed about 0.2 grams of the sample into a plastic beaker of 400 mL, then added 10 mL of concentrated HNO3 solution, and added about 5 mL of 40% hydrofluoric acid, then shook for about 30 minutes. Added 5 mL of 5% urea solution, and stirred with a plastic rod until no bubbles were generated. Added 10 mL of 15% potassium fluoride solution, 2–3 g of potassium chloride, and stirred until dissolved. It was then cooled to room temperature in cold water. The reaction products were filtered with a quantitative medium speed filter paper and the plastic funnel and the precipitates were washed with 10 mL of potassium nitrate-ethanol solution each time, a total of 2 times. The precipitates together with the filter paper were transferred to the original plastic beaker. Added 15 mL of saturated potassium chloride-ethanol solution and 5–6 drops of phenolphthalein. The residual acid was neutralized with 5% sodium hydroxide solution. After carefully scrubbing the inner wall of the beaker, the filter paper and the precipitates were stirred until a stable rose red color appeared. Then, 150 mL of boiling water was added, and 5 drops of phenolphthalein indicator were added additionally, and the solution was titrated with sodium hydroxide standard solution immediately until a steady reddish color appeared and no longer disappeared after agitation, which was taken as the end point. The silicon content is given, as a percentage by mass, by the formula:
where
C is the concentration, in moles per litre, of the sodium hydroxide standard volumetric solution; V is the volume, in millilitres, of the sodium hydroxide standard volumetric solution used for the titration; G is the mass, in grams, of the test portion.
Characterization
X-ray powder diffraction (XRD) patterns of the products were recorded on a Philips X’pert X-ray diffractometer with Cu Ka radiation (k = 1.54182 Ao). The microstructure was observed on a scanning electron microscope (KYKY-2800B SEM). The size distribution was measured using BT-2003 laser particle size analyzer.
Electrochemical Measurements
Electrochemical performance was tested using coin-type (CR 2016) cells with lithium foil as counter and reference electrodes. The working electrode was fabricated by coating a paste of C/Si composites and polyvinylidene fluoride (PVDF) binder (90:10 wt.%) on an copper foil collector. The electrode was dried at 110°C for 12 h in a vacuum oven under vacuum before assembly into a coin cell in an argon-filled glove box. The nonaqueous electrolyte used was 1 M LiPF6 dissolved in an ethylene carbonate (EC)-dimethyl carbonate (DMC)-diethyl carbonate (DEC) mixture (1:1:1, in wt.%). Galvanostatic cycling experiments of the cells were performed on a LAND CT2001A battery test system in the voltage range of 0.001–1.5 V versus Li+/Li at room temperature. The LiFePO4 cathode (capacity of ca. 162 mAh g–1, anode/cathode capacity ratio of ca. 1.15) was used for the full cell tests. The fabrication method of the full cell was the same as the half cell. Electrochemical performance was carried out within a potential window of 2.3–3.7 V at the current density of 0.1 A g–1.
RESULTS AND DISCUSSION
Chemical Composition of C/Si the Obtained Composites
A total of three C/Si composites with different proportions of carbon and silicon were synthesized. After chemical analyses, the chemical composition was as follows:
It can be seen from the above analyses that the reaction of potassiothermic reduction of silicon tetrachloride is complete at 600°C, and the decomposed silicon is mixed with the raw graphite anode materials. The preparation of the C/Si composites was realized.
Chemical composition of the obtained C/Si composites
XRD Pattern of the C/Si Composites
XRD patterns of 1#, 2# and 3# samples were measured. It can be seen from the XRD patterns that the crystal structure of the graphite during the reaction was not destroyed, and the silicon in the C/Si composites prepared has a good crystallinity. With the increase of the content of silicon in the C/Si composites, the XRD patterns have no change. Moreover, there are no other impurity peaks, indicating that no other impurities are formed during the reaction.
XRD pattern of the C/Si composites prepared.
SEM images of the graphite and the obtained C/Si composites: (a) SEM image of the raw material graphite, (b) SEM image of 1# sample, (c) SEM image of 2# sample, and (d) SEM image of 3# sample.
Analysis of the Morphology of Graphite Raw Materials and C/Si Composites
It can be seen from the SEM images that the graphite of the raw material has a particle size of about 10–20 micrometers, and the morphology is relatively regular, and the edges between the particles are relatively distinct. After compositing with the silicon, the particle sizes do not change much, but the morphology becomes complicated, and the boundaries between the particles are no longer clear. In Fig. 4. SEM-EDS images, some silicon particles latch onto the surface of the graphite particles, and some of the others exist separately. Comparing with the chemical analysis, the results of element contents have a considerable deviation because of the limitation of EDS element analysis. The chemical analysis method should be standard.
SEM-EDS images of the 2# sample.
Electrochemical Performance
The obtained three C/Si composites were tested for electrochemical performance. The current density of 0.3 A g–1 was used for constant current charge and discharge tests, and the voltage range of the tests was 0.005–1.5 V.
Fig. 5a–c are the charge and discharge curves for the 1st, 500th and 1000th cycles of 1#, 2# and 3# samples, respectively. Fig. 5d is the cycle performance curves for the three composites. It can be seen from Fig. 5a, b and c that the charge and discharge voltage of the 1# sample is lower than that of the 2# and 3# samples, mainly because the content of the 1# sample i s low, and the charge and discharge voltage is mainly contributed by carbon. It can be seen that under the same conditions of charging current, the charge and discharge voltage of silicon is about 0.1–0.2 V higher than that of carbon.
(a, b, c) Galvanostatic charge/discharge profiles of the 1st, 500th and 1000th cycles of the half cells for 1#, 2# and 3# samples. (d) Cyclic performance and Coulombic efficiency of the half cell for 1#, 2# and 3# samples. (e) Cyclic voltammetry (CV) curves for 1#, 2#, 3# samples and the graphite. (f) Galvanostatic charge/discharge profiles of the 1st, 500th and 1000th cycles of the full cells for 2# sample. (g) Cyclic performance and Coulombic efficiency of the full cell for 2# sample with the LiFePO 4 as cathode materials. (h, i) Electrochemical impedance spectroscopy (EIS) of the half cell for graphite and 1#, 2# and 3# samples before cycling and after 500 cycles. (j) The equivalent circuit of the half cell.
As can be seen from Fig. 5d, as the content of silicon in the C/Si composites increases, the initial capacity of the prepared C/Si composites increases, but the capacity fading rate becomes faster. The initial capacity of 1# sample is about 640 mAh g–1, and after 500 cycles, the capacity is reduced to 600 mAh g–1, which is 93.1% of the theoretical capacity, and after 1000 cycles, the capacity is reduced to 570 mAh g–1, which is 88.4% of the theoretical capacity. The initial capacity of 2# sample is about 960 mAh g–1, and after 500 cycles, the capacity is reduced to 880 mAh g–1, which is 89.7% of the theoretical capacity, and after 1000 cycles, the capacity is reduced to 820 mAh g–1, which is 83.6% of the theoretical capacity. The initial capacity of 3# sample is about 1260 mAh g–1, and after 500 cycles, the capacity is reduced to 860 mAh g–1, which is about 66.5% of the theoretical capacity, and after 1000 cycles, the capacity is reduced to 710 mAh g–1, which is about 54.9% of the theoretical capacity. This is mainly because as the content of silicon in the C/Si composites increases, the accumulation of silicon in the carbon layer becomes thicker, and the capacity fading is accelerated. 1# sample has the least content of silicon, and the capacity retention rate is the best, but the specific capacity is low. The 3# sample has the largest content of silicon, and the initial capacity is the highest, but the attenuation of capacity is rapid. After about 450 cycles, the specific capacity is less than 2#. Therefore, through the comprehensive comparison, 2# has the best content of silicon in the C/Si composites. Fig. 5d and Table 3 shows the Coulombic efficiency of the graphite and the C/Si composites. The initial Coulombic efficiencies (ICE) were 95.3%, 94.5%, 91.3% and 80.6% for the graphite, 1# sample, 2# sample and 3# sample respectively. After first three cycles, the ICEs achieved more than 99.5% for all samples, suggesting that the cells possessed good cycling reversibility. Fig. 5e is the cyclic voltammetry (CV) curves for the samples. The half-cells stabilized after 30 charge–discharge cycles at the current density of 0.3 A g–1 were scanned between 0.005 and 1.0 V vs. Li+/Li at a scanning rate of 0.01 mV s–1. In the charging scan, there are three distinct peaks around 0.21 V, 0.36 V and 0.51 V. Comparing with the published literature [22,23], the 0.51 V and 0.36 V peaks should correspond to the delithiation of the Li–Si alloy, while the 0.21 V peak should correspond to the delithiation of LixC6. In the discharging scan, the peak around 0.06 V should correspond to the lithiation of the graphite, while the 0.17 V peak should correspond to the combined lithiation of the graphite and the silicon. Fig. 5f and g show that the ICE and the first discharge capacity of the full cell are 89.1% and 114.27 mAh g–1, and after 1000 cycles, the discharge capacity is 98.05 mAh g–1, and the capacity retention rate is ca. 85.8%. Fig. 5h–j are Nyquist plots of the graphite and C/Si electrodes before cycling and after 500 cycles and their equivalent circuit. Among the simulated kinetic parameters, Re represents electrolyte resistance, Rf is the Ohmic resistance of the SEI film on the electrode, Rct is the charge transfer resistance related to the active material, while CPE shows the double layer capacitance, and W is the Warburg impedance for the diffusion of Li+ in the electrode. Before cycling as shown in Fig. 5h, with the increase of silicon contents, the diameter of the C/Si composites electrodes become larger and large, indicating that the Rct values become greater and greater. After 500 cycles in Fig. 5i, a new small arc appeared, maybe associated with the Ohmic resistance of the SEI film (Rf). The values of Re, Rf and Rct all increased after 500 cycles.
The Coulombic efficiency of the first three cycles
CONCLUSIONS
By incorporating silicon into the graphite by a simple one-step reaction, the specific capacity of the graphite anode can be greatly increased. The C/Si composites obtained by the potassiothermic reduction of silicon tetrachloride method has low impurity content, high purity and good electrochemical performance. However, the content of silicon added should not be too large, and the excessively large content of silicon leads to an increase in capacity fading. when the content of silicon in the C/Si composites is 19.0%, the theoretical specific capacity of the obtained C/Si composites is 981.33 mAh g–1, and after 1000 cycles of charge and discharge, the actual specific capacity can still reach 820 mAh g–1, which is about 83.6% of the theoretical specific capacity. In the full cell tests with LiFePO4 as the cathode, the ICE and the first discharge capacity are 89.1% and 114.27 mAh g–1, and after 1000 cycles, the discharge capacity is 98.05 mAh g–1, and the capacity retention rate is ca. 85.8%. Comparing with the simple mechanical mixing method [18,24] and organic compounds pyrolysis method [19–21], our C/Si composites synthesized via potassiothermic reduction have better cycling performance and capacity retention. This study has very great practical significance. Traditional graphite anode materials producers can significantly increase the price/performance ratio of their products through simple technical innovation.
Notes
ACKNOWLEDGEMENTS
This work was supported by the Doctor Foundation of Heze University (No. XY20BS01).