The n-type Si wafers, respectively, with electrical resistivity values of and used in this work were purchased from Hangzhou JingBo technology co., LTD. After being ball-milled at 400 rpm for 20 h (SFM-1, Hefei KeJing Materials Technology Co., LTD), Si wafers were refined into submicron Si particles. In our previous work,^[25] p-type Si particles were deposited with Ag via self-selective electroless deposition. Hence, we use the same method to prepare Ag-deposited n-type Si particles. After being cleaned by hydrofluoric acid (HF) solution with a volume ratio of 3%, 3-g Si particles were, respectively, reacted with 3-M HF and 17-mM silver nitrate (AgNO₃) in a 300-mL solution for 5 min. In the process, 200-mL H₂O was first added to disperse Si particles, and the remaining mixed solution was replenished 3 min later. After being filtrated and washed with H₂O, the Ag-deposited n-type Si particles were dried under vacuum at 40 °C for 4 h.

Scanning electron microscopy micrographs (SEM) were captured on JEOL JSM-6701 F at 15-kV accelerating voltage. The energy dispersive spectrum (EDS) was obtained from the energy dispersive spectrometer equipped with FE-SEM. Inductively coupled plasma atomic emission spectrometry (ICP-AES, PE Optima8000) was employed to measure the mass content of Ag. Laser diffraction instrument (BT-9300 H, Liaoning, China) was used to measure the sizes of Ag-deposited n-type Si particles. Powder x-ray diffraction (XRD) was obtained by Bruker D8 Advance x-ray diffractometer with using the Cu Kα radiation in steps of 0.04° s⁻¹. The electrical resistivity values of the four samples were measured on a semiconductor powder electrical resistivity tester (ST2722, Beijing, China).

The anode materials were slurry-cast on copper current collectors. Specifically, Ag-deposited Si particles, super-P and sodium alginate were ground in weight ratio of 7:1.5:1.5, then the as-prepared mixtures with 6-ml H₂O and 2-ml absolute ethanol (C₂H₅OH) were stirred to form slurries. All slurries were coated on copper current collectors, then dried under vacuum at 120 °C for 12 h. Finally, CR2025 coin cells containing as-prepared working electrode, diaphragm, electrolyte and lithium tablets as the reference electrode were assembled inside glove box filled with high-purity argon. The diaphragm was a Celgard 2400 sheet with a diameter of 19 mm. The mixture of 1-M LiPF₆/EC: DMC (V:V = 1:1) was used as the electrolyte.

The LAND battery cycler (LAND BT1-10, Shenzhen, China) was employed to measure the galvanostatic charge-discharge properties of the four samples. The potential voltage ranged from 0.01 V to 1.5 V. Electrochemical impedance spectrum (EIS) was recorded by a Princeton Applied Research spectrometer. The frequency ranged from 0.01 kHz to 100 kHz with an alternating current voltage of 10 mV. All the measurements were carried out at 25 °C.

The Si particles milled from n-type Si wafers, respectively, with an electrical resistivity of and are deposited with Ag in the same process. Figure 1(a) and 1(b) show SEM images of Ag-deposited n-type 1--Si particles (n1-Ag) and Ag-deposited n-type 0.001--Si particles (n0.001-Ag), respectively. The ellipsoidal particles with dozens of nanometer in size are attributed to Ag nanoparticles, while submicron Si particles exhibit irregular shapes. Obviously, Ag nanoparticles are evenly dispersed on the surfaces of Si particles. The insets in Figs. 1(a) and 1(b) show EDS and ICP results of n1-Ag and n0.001-Ag, respectively. It can be found in the two figure that the mass percent values of Ag in the surface are almost the same. Moreover, there is no significant difference between the overall Ag content of n1-Ag and that of n0.001-Ag, which can be obtained from the ICP results. The XRD patterns of n1-Ag and n0.001-Ag are shown in Fig. 1(c). The characteristic peaks of Si and Ag are clearly identified. In addition, there are no obvious difference between the full widths at half maximum of Ag in n1-Ag and n0.001-Ag, and so is the scenario of Si. It indicates that the crystal structures of Si and Ag in the n1-Ag and n0.001-Ag are the same.

Fig. 1.

	Figure Option ViewDownloadNew Window
	Fig. 1. (color online) SEM images of (a) n1-Ag and (b) n0.001-Ag, (c) x-ray diffraction patterns and (d) particle size distribution curves of n1-Ag and n0.001-Ag.

Figure 1(d) presents that the particle size dispersion curves of n1-Ag and n0.001-Ag. The particle size distribution curves of these two samples basically overlap, which indicates that their particle size is basically the same. Moreover, the median-particle-size D50 of n1-Ag and n0.001-Ag are and , respectively. Accordingly, the n1-Ag and n0.001-Ag present similar morphologies, the same crystal structure and undifferentiated size distribution.

Figure 2 shows the structural diagram and energy band diagram of n1 and n0.001 before and after being deposited with Ag. Ag, as a metal material, has a work function fixed at 4.26 eV. As a semiconductor material, the work function of Si varies with the type and concentration of doping. In this work, the doping concentration of n-type Si with electrical resistivity of and are and , respectively. It is calculated that the work function of n0.001 (W_n0.001) is 4.08 eV, which is 0.18 eV lower than that of Ag.^[22] Thus, the energy band of n0.001 is bent downward after deposited with Ag, which generates a Schottky barrier.^[22–24] The Schottky barrier causes the contact resistance to rise. On the contrary, the work function of n1 (W_n1) is 4.31 eV, which is 0.05 eV higher than that of Ag. After being deposited with Ag, the energy band of n1 is bent upward to reduce the contact resistance. Therefore, the resistivity of n1-Ag is lower than that of n0.001-Ag.

Fig. 2.

	Figure Option ViewDownloadNew Window
	Fig. 2. (color online) Structural diagram and energy band diagram of n1 and n0.001 before and after being deposited with Ag.

The semiconductor powder electrical resistivity tester is used to test the powder resistivity of Si particles and Ag-deposited Si particles in order to characterize their resistivity changes. Powder resistivity values of different samples are shown in Table 1. It can be observed that the powder resistivity of n1 and n0.001 are both several hundred thousand times that of the corresponding Si wafer. This is attributed to the fact that the natural oxide layer on the surface of Si particles increases the contact resistance among them. The powder resistivity values of n1-Ag and n0.001-Ag are and , respectively, which are lower than that of corresponding Si particles. However, it is worth noting that the powder resistivity of n1-Ag is lower than half that of n0.001-Ag, while that of n1 is 210 times higher than that of n0.001. Specifically, the powder resistivity of n1-Ag is times lower than that of n1, while that of n0.001-Ag is only times lower than that of n0.001. The theoretical analysis mentioned above indicates that the matching of work function between n1 and Ag leads to these results.

Table 1.

Table 1.

Table 1. Powder resistivity of different samples. . Sample n1 n0.001 n1-Ag n0.001-Ag Resistivity

Table 1. Powder resistivity of different samples. .

The EIS of the four electrodes are displayed in Fig. 3(a). It can be found that the intercept on x axis in the high frequency region of n1-Ag is smallest. It indicates that n1-Ag has the lowest impedance. It is consistent with the theoretical analysis mentioned above. Figure 3(b) shows the voltage profiles corresponding to the first cycle of the four electrodes at a current density of . The discharge specific capacities of n1, n0.001, n1-Ag, and n0.001-Ag in the first cycle are , , , and , respectively. Due to the presence of Ag, the discharge capacity of Ag-deposited Si particles is lower than that of the corresponding Si particles without Ag, respectively.

Fig. 3.

	Figure Option ViewDownloadNew Window
	Fig. 3. (color online) (a) EISs of the four electrodes before cycling, (b) the first cycle charge/discharge profiles, (c) rate performance and (d) cyclabilities of four electrodes.

Figure 3(c) shows the rate performances of the four types of electrodes at current densities of , , , , , and . Except the first five cycles, the discharge specific capacity of n1-Ag is always highest. In particular, the discharge specific capacity of n1-Ag is at a current density of , which is 40% higher than that of n0.001-Ag. The cycle performances of the four types of electrodes are shown in Fig. 3(d). After 100 cycles, the discharge specific capacities of n1, n0.001, n1-Ag, and n0.001-Ag are , , , and , respectively. Obviously, the discharge specific capacity of n0.001 is higher than that of n1. This is consistent with our previous result.^[26] Dramatically, the discharge specific capacity of n0.001-Ag is lower than that of n1-Ag. This can be attributed to the matching of work function between n1 and Ag, which improves the electrical connection between electrode and active material.