Introduction

Approximately 150 million tons of rice husk (RH) are generated annually and globally as agricultural waste1. RH is composed of approximately 80 mass% organic material (lignin, cellulose, hemicellulose, and others), and the rest are mainly silicon oxides (SiOx, 0 < x < 2)2. Rice, a silicicolous plant, adsorbs Si(OH)4 from the soil for growth and accumulates SiOx in its husk3. RH is not preferred as a fuel because it produces a large amount of SiOx-derived ash during combustion. Thus, the utilization of RH has been limited to agricultural applications such as fertilizer additives, stockbreeding mats, and bed soil. On the other hand, RH, being intrinsically a composite of organic material and inorganic SiOx itself, can be easily transformed into a C/SiOx composite through simple heat treatment under inert conditions (carbonization).

Lithium-ion batteries (LIBs) have been used as power sources in various industrial applications, such as smartphones, industrial robots, electric vehicles, and grid-scale electricity storage systems. Graphite, hard carbon, Si, and SiOx have been used as anodic active materials (AMs) in LIBs4,5,6,7,8,9,10. Anodic AMs operate using several Li storage mechanisms (insertion, conversion, and alloying reactions)8,11,12,13,14. Intercalation compounds, such as carbon materials, follow the intercalation mechanism, which facilitates Li-ion insertion (reduction) and extraction (oxidation)11. This mechanism provides a good conductive pathway, smooth Li-ion transport, and strong robustness during cyclic Li-ion insertion and extraction. Alloying compounds, such as Si and Sn, follow the alloying mechanism when they react with Li, which can lead to dramatically higher specific capacities for Li storage than those for the insertion mechanism14. High specific capacities are is also obtainable by the conversion mechanism associated with the reactions of transition metal oxides and Li. However, large irreversible capacities and positive anode potentials are induced by this mechanism compared to the other mechanisms.

Low electrode potential, high specific capacity for Li-ion uptake and release, good cycle stability, low toxicity, low cost, and natural abundance are required for anodic AMs. Graphite, which follows the insertion mechanism, is still mainly used in commercial LIBs because of its low and flat potential profile, low irreversible capacity, low cost, and accumulated experience. However, it has drawbacks, including a limited specific capacity of 372 mAh g−1, small margin for Li deposition or dendrite formation, and moderate cycle life11,15. Si, which follows the alloying mechanism, is anticipated to supersede graphite because of its very high specific capacity (4200 mAh g−1 at full Li alloying)14. However, it undergoes a large volume expansion during Li-ion uptake and release (~ 280%)16, causing the cracking of Si particles and thus, repeated formation of a solid-electrolyte interphase (SEI). Repeated SEI formation consumes Li ions in the electrolyte and increases the irreversible capacity of the Si anode. Moreover, cracking can induce the delamination of Si particles from the electrode, leading to capacity fading of the cell. Therefore, the sole use of Si in LIB anodes is still rare for commercial LIBs. However, the use of SiOx has been an alternative countermeasure to increase the anode capacity of cells. High specific capacities of 1965 and 2043 mAh g−1 were obtained for SiO2 and SiO, respectively7,17. The reduction reaction of SiOx with Li ions can produce lithium silicate (LixSiOy) and lithium oxide (Li2O), leading to a smaller volume expansion at succeeding cycles of Li-ion uptake and release and thus, higher cycle stability17. However, SiOx has a lower electrical conductivity than C and Si, requiring a high-level conductive agent (e.g., carbon black) to fabricate the anode. Thus, C/SiOx composites have attracted much attention as anodic AMs for LIBs18,19,20,21.

For sustainable development goals, biomass-derived materials are viable options for next-generation LIB anodes that require high environmental compatibility. As mentioned above, C/SiOx composites can be readily produced from RH via simple carbonization. RH-derived C/SiOx composites have been investigated for application as LIB anodes. Ju et al. stated that RH-derived C/SiOx exhibits a large irreversible capacity of 46.7% Coulombic efficiency (CE) at the initial Li-ion insertion and extraction22. Kumagai et al. also reported ~ 53% CE at the initial Li-ion uptake and release17. The large irreversible capacity of RH-derived C/SiOx has been demonstrated elsewhere15,20,21,22,23. In addition, the Li-ion insertion and extraction properties of carbon materials produced by removing almost all SiOx from carbonized RH were also evaluated24,25. Although SiOx undergoes reduction reactions with Li ions and thereby increases the irreversible capacity, the CE of the carbon material is not high (~ 60%)15,18. SiOx in RH acts as a template to produce micro- and mesoporous structures in the resultant carbon material26. The passivation of Li ions in the produced pores, as well as SEI formation on the increased surface area, is responsible for the large irreversible capacity20,23. To overcome the large irreversible capacity of RH-derived C/SiOx, a pre-lithiation process that provides sufficient Li ions to the anodic AMs can be a realistic choice27,28,29,30,31,32. Up to now, the Li-ion insertion and extraction properties of RH-derived C/SiOx or C have been evaluated in the half-cell configuration where infinite Li ions can be provided by the Li metal electrode (counter and reference electrodes). However, for actual LIB full-cells, the large irreversible capacity of the anode has a decisive effect on the charge–discharge performance because a significant amount of Li ions can be consumed within and on the anode, and additional Li ions cannot be supplied from the outside. Although several studies have focused on the feasibility of RH-derived C/SiOx as a LIB anodic AM, its electrochemical performance in a full-cell configuration has not been reported thus far. The role of SiOx in LIB full-cell performance is also unknown.

In the present study, C/SiOx was produced by heating RH under N2 and then immersing the carbonized RH in NaOH solution. By adjusting the immersion temperature and time, C/SiOx composites with different SiOx contents were prepared as LIB anodic AMs. The physical properties of the prepared samples were evaluated by thermogravimetric analysis, particle size analysis, and N2 adsorption–desorption porosimetry. CR2032-type half-cells and full-cells were assembled to evaluate the Li-ion insertion and extraction properties of C/SiOx. Pre-lithiation was performed prior to the LIB full-cell assembly. Full-cells with a pre-lithiated or pristine C/SiOx anode and Li(Ni0.5Co0.2Mn0.3)O2 (NCM) cathode were assembled, and their rate and cycle performance were evaluated. Finally, the role of SiOx in the C/SiOx anode in the rate and cycle performance of the assembled full-cells was discussed. To the best of our knowledge, the performance of LIB full-cells with pre-lithiated C/SiOx anodes has been evaluated for the first time in this study. Revealing the roles of the pre-lithiation process and SiOx removal in the anode on the performance of LIB full-cells is scientifically novel and industrially and environmentally significant. The knowledge provided in this study is expected to promote the development of eco-friendly LIBs and the effective use of agricultural waste.

Experimental methods

Preparation of C/SiOx AMs and their characterization

Raw RH (Akitakomachi rice) was provided by a rice farmer in Senboku City, Akita Prefecture, Japan with his permission and used as feedstock for RH-derived C/SiOx AMs. The raw RH was dried at 100 °C for 10 h to determine the initial mass. Pre-carbonization was conducted in a tubular furnace by heating the raw RH at 600 °C for 1 h under N2 gas flow at 1 L min−1. The pre-carbonized RH was then immersed in a 1 mol L−1 NaOH solution under the following two conditions: (A) immersion temperature: 25 °C, immersion time: 22 h for half-elimination of SiOx and (B) immersion temperature: 80 °C, immersion time: 10 h for complete elimination of SiOx. The mass ratio of pre-carbonized RH to the solution was 1:20. The pre-carbonized RH immersed in NaOH solution was rinsed with distilled water until the pH of the used rinse water was reduced to < 9. For the pre-carbonized RH that was not immersed in NaOH solution, the rinse consisting of distilled water was also prepared. All rinsed samples were dried at 100 °C for 10 h. For the main carbonization, they were heated to 1000 °C, and the temperature was maintained for 1 h in a similar manner to the pre-carbonization. Grains of the carbonized RH were pulverized using a planetary ball-milling machine with a zirconia vessel and balls (P-6, Fritch Japan Co., Ltd., Japan) at 400 rpm for 20 min, finally providing C/SiOx AMs with different SiOx contents.

The production mass yield was calculated based on the masses of the dried raw RH and prepared C/SiOx AMs. The average particle diameters of the C/SiOx AMs were measured using a laser diffraction particle size analyzer (SALD-200 V, Shimadzu Corp., Japan). The specific surface areas of the C/SiOx AMs were measured using a gas adsorption analyzer (Autosorb-3B, Quantachrome Instruments Inc., USA). The Brunauer–Emmett–Teller specific surface area was calculated from the obtained nitrogen adsorption–desorption isotherms at − 196 °C. The total pore volume was measured using the volume of adsorbed nitrogen at a relative pressure close to 1 as reference. A thermogravimetry system (Thermo Plus EVO TG8120, Rigaku Corp., Japan) was used to heat the C/SiOx AMs up to 850 °C under air flow (200 mL min−1). The SiOx content was calculated from the mass of C/SiOx at 140 and 850 °C. The physical properties of the three types of C/SiOx AMs are summarized in Table 1. The C/SiOx AMs are denoted as C/SiOx-45, C/SiOx-24, and C/SiOx-5 based on the SiOx content. The production mass yield decreased with the SiOx content and was reduced by ~ 14% owing to complete SiOx removal. The particle sizes of the C/SiOx AMs were adjusted to 4–5 µm. The removal of SiOx increased the specific surface area and total pore volume, which is related to the decrease in the production mass yield.

Table 1 Physical properties of the C/SiOx active materials.
Full size table

Electrode fabrication

To fabricate the anode, 80 mass% C/SiOx, 10 mass% conductive agent (acetylene black, Denka Co., Ltd., Japan), 7.5% sodium carboxymethyl cellulose (Cellogen 7A, DSK Co., Ltd., Japan), and 2.5 mass% styrene–butadiene rubber (TRD2001, JSR Corp., Japan) were dispersed in distilled water and mixed using a planetary centrifugal mixer (AR-100, Thinky Corp., Japan). The uniformly dispersed slurry was coated onto a copper foil (no surface treatment, t20 µm, Hohsen Corp., Japan) using an applicator and dried in an oven at 100 °C for > 6 h. The cathode for the full-cell assembly was also fabricated. A ternary lithium transition metal oxide (Li(Ni0.5Co0.2Mn0.3)O2, NCM, Beijing Easpring Material Technology Co., Ltd., China), acetylene black, and polyvinyl difluoride (KF polymer F #9130, Kureha Corp., Japan) were mixed at a mass ratio of 80:10:10 in N-methylpyrrolidone (Tokyo Chemical Industry Co., Ltd., Japan). The cathode slurry was coated onto an aluminum foil (no surface treatment, t20 µm, Hohsen Corp., Japan) and dried in an oven at 100 °C for > 6 h. Both dried electrodes were punched out into ϕ15-mm disks and dried at 140 °C under vacuum for 3 h. The loading masses of the AMs in the cathodes and anodes were respectively 5.57–5.61 and 2.15–2.31 mg cm−2; hence the electrode loading masses were 6.96–7.01 mg cm−2 for the NCM cathode and 2.69–2.89 mg cm−2 for the C/SiOx anode. The coating thicknesses of the cathode and anode layers on the current collectors were 42–44 and 31–43 µm, respectively.

Half-cell assembly and pre-lithiation of the C/SiOx electrodes

The half-cell consisting of the C/SiOx electrode and Li metal foil (ϕ15 mm, t0.2 mm, Honjo Metal Co., Ltd., Japan) was assembled using a CR2032 coin cell (Hohsen Corp., Japan) in an argon-filled glove box. The electrolyte used was 1 mol L−1 LiPF6 dispersed in 1:1 v/v% ethylene carbonate/diethyl carbonate (Kishida Chemical Co., Ltd., Japan). The two electrodes separated by a polypropylene separator (ϕ18 mm, 2500, Celgard LLC, USA) were installed in the coin cell after dipping the electrodes and separator in the electrolyte for ~ 5 s. A Li-ion insertion-extraction test was performed on the coin-type half-cells using a battery charge–discharge system (HJ1020mSD8, Hokuto Denko Corp., Japan). The electrode potential range was 0–2.5 V vs. Li+/Li, and the current density was 20 mA gAM−1, where gAM is the unit of the mass of the C/SiOx AM in the electrode. The Li-ion insertion-extraction process was repeated five times. The specific capacities of the C/SiOx AMs were determined using the half-cell test results as the reference. The cycling stabilities of the C/SiOx AMs were then evaluated by repeated (100-times) Li-ion insertion-extraction at a current density of 200 mA gAM−1 in a similar potential range.

Assembly of the full-cells with pre-lithiated C/SiOx electrode and their evaluation

The specific capacity of the employed NCM AM was 150 mA g−1 based on the previous test results obtained using a half-cell configuration28. According to the specific capacity of the C/SiOx AMs determined in the coin-type half-cell tests, the capacity ratio of the anode to the cathode in the full-cells was adjusted to 1.0 − 1.1. The pre-lithiation of the C/SiOx electrode and NCM + C/SiOx full-cell assembly is illustrated in Fig. 1. To prepare the pre-lithiated anode, a half-cell with the C/SiOx electrode was separately assembled using an SUS304 stainless steel flat-type cell (HS Flat cell, Hosen Corp., Japan). The flat-type half-cell consisting of the C/SiOx electrode, the above-mentioned Li metal foil, separator (ϕ23 mm), and electrolyte (1 mL) was assembled. The C/SiOx electrode was pre-lithiated in an assembled flat-type half-cell. The potential of the C/SiOx electrode decreased at a current density of 40 mA gAM−1, delivering Li ions to the C/SiOx AM. The pre-lithiation sequence was stopped when the electrode potential reached 0 V vs. Li+/Li. Following pre-lithiation, the flat-type half-cell was introduced into an argon-filled glove box. The pre-lithiated C/SiOx electrode was carefully extracted from the flat-type half-cell. Coin-type full-cells were then fabricated using the NCM cathode and pre-lithiated or unlithiated C/SiOx anode in a similar manner to the coin-type half-cells.

Figure 1
figure 1

Pre-lithiation of the C/SiOx anode and Li(Ni0.5Co0.2Mn0.3)O2 (NCM) + C/SiOx full-cell assembly.

Full size image

The C-rate was employed to represent the current density of the full-cells. 1 C is defined as 150 mA gCAM−1, where gCAM is the unit of the mass of the cathodic AM. The cell voltage range used for the charge–discharge sequences of the full-cells was 2.5–4.2 V. Charge–discharge cycling at 0.1 C was first performed three times for cell stabilization. Following cell stabilization, the rate performance of the full-cells was evaluated, during which the C-rate was increased stepwise from 0.1 to 10 C, and then decreased to 0.1 C. The charge–discharge process was repeated five times at each C-rate. The cycle performance of the full-cells was then evaluated by repeating the charge–discharge cycles at 2 C. The full-cell performance was represented by the cathode specific capacity (time integral of the cell current divided by the mass of cathode AM in units of mAh gAM−1).

Results and discussion

Li-ion insertion and extraction properties of C/SiOx AMs

Figure 2 shows the initial Li-ion insertion and extraction properties of the C/SiOx AMs at 20 mA gAM−1, and their cycling stabilities at 200 mA gAM−1 in the half-cell configuration. In the first cycle, C/SiOx-5 exhibited the highest Li-ion insertion specific capacity of 752 mAh gAM−1. A comparative Li-ion insertion specific capacity (~ 650 mAh g AM−1) was observed for C/SiOx-45 and C/SiOx-24. All C/SiOx AMs displayed a Li-ion extraction specific capacity of < 410 mAh gAM−1, which was much lower than those for Li-ion insertion. The initial CEs of C/SiOx-45, C/SiOx-24, and C/SiOx-5 were 63.0%, 54.1%, and 50.2%, respectively, indicating that the initial CE decreases with decreasing SiOx content. The CEs increased with the number of cycles. In the fifth cycle, C/SiOx-45 had the highest Li-ion extraction specific capacity (420 mAh gAM−1), while C/SiOx-24 had the lowest one (298 mAh gAM−1). C/SiOx-5 displayed a maximum specific capacity of 380 mAh gAM−1. The apparent specific capacity of the C/SiOx AMs was determined as 400 mAh gAM−1. These results demonstrate that SiOx is beneficial for increasing the specific capacity of C/SiOx AMs. On the other hand, partial removal of SiOx is detrimental to maintaining the specific capacity. Then, the cycling stabilities of C/SiOx AMs were evaluated at 200 mA gAM−1. The specific capacities for Li-ion extraction from all C/SiOx AMs decrease significantly within the initial 10 cycles in the half-cell configuration where ample Li-ions are delivered, after which they plateau during the remaining 100 cycles.

Figure 2
figure 2

Initial Li-ion insertion and extraction properties of (a) C/SiOx-45, (b) C/SiOx-24, and (c) C/SiOx-5 in a coin-type half-cell configuration at 20 mA gAM−1, and (d) their cycling stabilities at 200 mA gAM−1 in the half-cell configuration in the 0–2.5 V (vs. Li+/Li) electrode potential range.

Full size image

Pre-lithiation of the C/SiOx electrodes

The C/SiOx AMs were pre-lithiated prior to full-cell assembly. The Li-ion insertion profiles of the C/SiOx AMs in the flat-type half-cells are shown in Fig. 3. C/SiOx-45 and C/SiOx-24 displayed a similar Li-ion insertion specific capacity of ~ 680 mAh gAM−1, while C/SiOx-5 had a higher specific capacity of 832 mAh gAM−1. The C/SiOx AMs exhibited different profiles at the first Li-ion insertion, and their potential levels decreased with increasing SiOx content. The Li-ion insertion specific capacities of the C/SiOx AMs obtained here were higher than those obtained using the coin-type half-cell, which can be attributed to the ample amount of electrolyte (1 mL) covering the electrodes.

Figure 3
figure 3

Li-ion insertion profiles of the C/SiOx active materials during pre-lithiation.

Full size image

Performance of the NCM + C/SiOx full-cells

Full-cells with the NCM cathode and pre-lithiated C/SiOx anode were fabricated. A full-cell incorporating the unlithiated C/SiOx-45 anode (NCM + C/SiOx-45 without pre-lithiation) was also fabricated and evaluated. The charge–discharge specific capacities and CEs of the full-cells during the initial three cycles at 0.1 C are summarized in Table 2. In the first cycle, a large difference between the charge and discharge specific capacities (176 and 17 mAh gCAM−1, respectively) and a very low CE of 9.7% were observed for NCM + C/SiOx-45 without pre-lithiation. Even after the succeeding cycles, the discharge specific capacity was ~ 20 mAh gCAM−1. By pre-lithiating the C/SiOx anodes, the CEs of the full-cells significantly increased up to ~ 85% in the first cycle. All full-cells with pre-lithiated anodes (pre-lithiated full-cells) exhibited stable charge–discharge profiles and very high CEs during the initial three cycles. In the 2nd cycle, the discharge specific capacities of the pre-lithiated full-cells reached the apparent specific capacity of the NCM cathode (150 mAh gCAM−1), indicating negligible loss of Li ions from the cathode.

Table 2 Charge (CC) and discharge (CD) specific capacities and Coulombic efficiencies (CE) of the Li(Ni0.5Co0.2Mn0.3)O2 (NCM) + C/SiOx full-cells at the initial 3 cycles at 0.1 C.
Full size table

Following the initial three cycles, the rate performance of all full-cells was evaluated at a current density of 0.1–10 C and is shown in Fig. 4. Except for NCM + C/SiOx-45 without pre-lithiation, the full-cells displayed similar levels of discharge specific capacity up to 2 C. At higher current densities (5 and 10 C), a higher SiOx content led to a higher discharge specific capacity. The discharge specific capacities of NCM + C/SiOx-45, NCM + C/SiOx-24, and NCM + C/SiOx-5 at 10 C were 60, 43, and 31 mAh gCAM−1, respectively. The CEs of the pre-lithiated full-cells were ~ 100%, except during the first cycle at each C-rate. The discharge specific capacity of NCM + C/SiOx-45 without pre-lithiation was much lower than those of the pre-lithiated full-cells throughout the rate test. The cell voltage-specific capacity profiles of NCM + C/SiOx-45 and NCM + C/SiOx-24 exhibited slope-like variations, while NCM + C/SiOx-5 maintained a higher cell voltage (allowing narrow plateau regions) than the others at low current densities (up to 2 C). Maintaining a high cell voltage increases the density of the energy delivered during the charge and discharge processes. At high current densities (> 2 C), the residual SiOx in the anode was effective in sustaining the cell capacity.

Figure 4
figure 4

Rate performance of the Li(Ni0.5Co0.2Mn0.3)O2 (NCM) + C/SiOx full-cells. (a) Discharge specific capacity and (b) Coulombic efficiency at different C-rates. Cell voltage-specific capacity profiles of the full-cells: (c) NCM + C/SiOx-45, (d) NCM + C/SiOx-24, (e) NCM + C/SiOx-5, and (f) NCM + C/SiOx-45 without pre-lithiation. The profiles of the 3rd cycle at each C-rate are shown. The cell voltage range was 2.5–4.2 V.

Full size image

The cycling performance of the full-cells at 2 C is shown in Fig. 5. The highest discharge specific capacity was displayed by NCM + C/SiOx-45 in the 1st cycle and NCM + C/SiOx-5 in the final cycle. All full-cells showed a gradual decrease in capacity with the number of cycles. The capacity retentions of NCM + C/SiOx-45, NCM + C/SiOx-24, and NCM + C/SiOx-5 after 800 cycles were 47.5%, 61.7%, and 60.7%, respectively. The presence of SiOx in the anode led to a ~ 13% difference in the retention of the cell specific capacity. The largest decrease was observed for NCM + SiOx-45, which has the highest level of SiOx. The discharge specific capacity of NCM + C/SiOx-45 without pre-lithiation was negligible throughout the 800 cycles, suggesting that pre-lithiation is essential for RH-derived C/SiOx AMs. Cycle aging was alleviated when the SiOx in the anode was removed completely or partially. However, NCM + C/SiOx-24 showed an unstable variation in the discharge specific capacity up to the ~ 400th cycle.

Figure 5
figure 5

Cycling performance of the Li(Ni0.5Co0.2Mn0.3)O2 (NCM) + C/SiOx full-cells. The cell voltage range was 2.5–4.2 V, and the charge–discharge current density was 2 C (300 mA gCAM−1).

Full size image

Effect of SiOx in the anode on the full-cell performance

The number of oxygen atoms (x) in SiOx in the RH carbonized at 1000 °C was found to be close to 217. Under this assumption, the SiOx in C/SiOx AMs can react with Li ions according to the following chemical formulae [Eqs. (1) to (4)]21:

Reversible reactions (Li-ion insertion and extraction)

$$5{\mathrm{SiO}}_{2}+4{\mathrm{Li}}^{+}+4{\mathrm{e}}^{-}\leftrightarrow 2{\mathrm{Li}}_{2}{\mathrm{Si}}_{2}{\mathrm{O}}_{5}+\mathrm{Si}$$
(1)
$$\mathrm{Si}+{\mathrm{xLi}}^{+}+{\mathrm{xe}}^{-}\leftrightarrow {\mathrm{Li}}_{\mathrm{x}}\mathrm{Si}.$$
(2)

Irreversible reactions (Li-ion insertion only)

$${\mathrm{SiO}}_{2}+4{\mathrm{Li}}^{+}+4{\mathrm{e}}^{-}\to 2{\mathrm{Li}}_{2}\mathrm{O}+\mathrm{Si}$$
(3)
$$2{\mathrm{SiO}}_{2}+4{\mathrm{Li}}^{+}+4{\mathrm{e}}^{-}\to {\mathrm{Li}}_{4}{\mathrm{SiO}}_{4}+\mathrm{Si}.$$
(4)

In the half-cell configuration, the reversible (Li-ion extraction) capacities of C/SiOx-45 and C/SiOx-5 were 420 and 380 mAh gAM−1, respectively. These results indicate that the specific capacities of C and SiOx were 375 and 475 mAh gAM−1, respectively. Based on the specific capacity of SiO2 (1965 mAh g−1)7, the irreversible reactions in Eqs. (3) and (4) dominate. However, approximately one-quarter of SiOx was responsible for the reversible specific capacity for Li-ion insertion and extraction via Eqs. (1) and (2). On the other hand, C/SiOx-24 had the lowest specific capacity of 298 mAh gAM−1. Sodium silicates such as Na2SiO3 are likely produced during SiOx leaching, i.e., immersion in NaOH solution20. They are inactive for Li-ion uptake and release, which reduced the specific capacity of C/SiOx-24. C/SiOx AMs with a lower SiOx content exhibited a lower initial CE. The removal of SiOx from C/SiOx AMs increased the specific surface area (44 m2 g−1 for C/SiOx-45 and 193 m2 g−1 for C/SiOx-5) because micro- and mesopores were produced at the trace of SiOx26. As pores develop in RH-derived anodic AMs, more Li ions can be passivated within the pores20,23. Thus, the lower the SiOx content, the lower the initial CE.

The physical properties of the C/SiOx AMs affected the rate and cycle performance of the NCM + C/SiOx full-cells. Pre-lithiation was demonstrated to be required for LIBs with RH-derived anodic AMs. Owing to pre-lithiation, all full-cells displayed a high discharge specific capacity of ~ 150 mAh gCAM−1 at 0.1 C. At the increased C-rates of 5 and 10 C, the discharge specific capacity was more retained when C/SiOx-45 was used as the anode, indicating that the time response of the reversible reactions of SiOx with Li ions [Eqs. (1) and (2)] is higher than that of the remaining C component. During the cycle test, the full-cell with the higher SiOx content could not retain the initial cell specific capacity, and unstable capacity variation was observed for NCM + C/SiOx-24. Although the SiOx component in C/SiOx AM provided a higher time rate for Li-ion insertion and extraction, the uptake and release of Li ions in SiOx [Eqs. (1) and (2)] seem to cause a large volume expansion, thereby leading to structural decay in the anode. The unstable variation of NCM + C/SiOx-24 could be explained by the formation of inactive sodium silicates during SiOx leaching.

The textural and electrochemical properties of RH-derived AMs intended for use in LIB anodes are summarized in Table 3. The reversible specific capacities of RH-derived AMs produced in the present study are lower than those reported in the relevant literature. However, C/SiOx-45, produced through two-step carbonization, exhibited the highest CE (63.0%) among all AMs, and the CEs of C/SiOx-24 and C/SiOx-5 are also high compared to those of the other AMs. A trade-off relationship between specific capacity and CE seems to be closely related to porosity, which is represented by the specific surface area or the total pore volume of the AM. Pre-lithiation was shown to be essential for the RH-derived AMs produced here when employed in a LIB full-cell, suggesting that pre-lithiation is necessary for all anodic AMs produced from RH for stable operation in LIB full-cells.

Table 3 Comparing the textural and electrochemical properties of RH-derived AMs intended for the in LIB anodes.
Full size table

In the present study, the partial removal of SiOx using the NaOH solution likely produced a sodium silicate byproduct, which is inactive for Li-ion uptake and release. This byproduct formation rather destabilized the cycle performance of the full-cell. It was shown that the specific capacity of the SiOx component was 100 mAh gAM−1 higher than that of the C component in the RH-derived anodic AM. In addition, the time response of the SiOx component for Li-ion uptake and release was also higher, thereby delivering the highest rate performance to the NCM + C/SiOx full-cells. However, the SiOx component reduced the specific capacity retention of the full-cells during the cycle test. Judging from the minor difference in the specific capacity retention (~ 13% in 800 cycles), significant effort to completely remove SiOx, excellent rate performance provided by SiOx, and formation of sodium silicate byproduct, it is not strongly necessary to remove SiOx from the RH-derived AM intended for LIB anodes. The important task of the present study was to develop sustainable and eco-friendly LIB anode materials. In this sense, it is preferrable to preserve SiOx to avoid using a strong alkaline solution for SiOx leaching and strong acid for neutralization. The high production yield due to SiOx preservation (~ 14 mass% higher than that for SiOx removal) is also beneficial for industrial applications.

Conclusions

RH-derived C/SiOx AMs with 45, 24, and 5 mass% SiOx were prepared by heat treatment and immersion in a NaOH solution. C/SiOx AMs that were pre-carbonized at 600 °C, then subjected to SiOx leaching, and finally carbonized at 1000 °C were evaluated as Li-ion battery anodes in half-cell and full-cell configurations. The specific capacities of the C and SiOx components were 375 and 475 mAh g−1, respectively, indicating that only about one-quarter of SiOx was responsible for the reversible specific capacity for Li-ion insertion and extraction. C/SiOx AMs with a lower SiOx content exhibited a lower initial CE for Li-ion insertion and extraction because of Li passivation within the pores developed at the trace of the removed SiOx. Full-cells with a NCM cathode and C/SiOx anode were assembled. Owing to the low initial CE (< 65%), pre-lithiation was essential to attain a stable anodic operation for the C/SiOx AMs. All full-cells exhibited a high initial CE (~ 85%) and high discharge specific capacity (~ 150 mAh gCAM−1) at 0.1 C. At the increased C-rates of 5 and 10 C, a higher SiOx content in the full-cell led to a higher cell specific capacity retention, revealing that the reversible reactions of SiOx with Li ions are faster than those of the C component. The full-cell with C/SiOx-45 exhibited the largest decrease in cell specific capacity (47.5% retention) during the cycle test, while the other full-cells exhibited ~ 60% retention. All results indicated that the SiOx component in the RH-derived AMs enhances the specific capacity and rate performance, while inducing structural decay owing to the volume expansion of SiOx. Considering the higher production yield, enhanced specific capacity and rate stability due to SiOx, acceptable cycle performance degradation, and operational effort, cost, and eco-friendliness of SiOx removal, the preservation of SiOx is recommended when carbonized RHs are utilized as LIB anodes.