Low-cost porous silicon material based on fenton process assisted chemical etching and its application as anode for lithium-ion battery

Nanostructured silicon-based materials are considered as one of the most promising alternatives for anode electrodes of lithium-ion batteries. Nevertheless, the relatively high cost and complexity of most present processes make them difficult to scale up. In this manuscript, we would like to introduce a porous anode prepared by Fenton reaction-assisted chemical etching of low-cost ferrosilicon in the mixed solution of hydrogen fluoride and hydrogen peroxide. The experimental results show that hierarchical structures containing both mesopores and micropores can be formed by this convenient room-temperature etching process and its electrochemical performance as the anode is comparable to that of commercial product. Furthermore, the low cost of ferrosilicon ($1200~1500/ton) compared with the silicon powder of industrial grade (~$3000/ton) traditionally used in Li-ion battery anode preparation also renders this method potential and feasible for large-scale production of Si-based anode materials.


Introduction
With the increasing demand for energy density of lithium-ion batteries, silicon-based composites have been used as anodes.The reversible intercalation process of lithium ions in the silicon frameworks determines the available capacity of the anode.It is well known that silicon (Si) will change from crystalline to amorphous phase during the process of intercalation, forming Li 15 Si 4 compound with a corresponding specific capacity of 3579 mAh/g, which is one of the best options to partially replace the graphite anode currently used [1].However, the formation of Si-Li alloy by intercalation of lithium ions in a silicon framework will lead to anisotropic volume expansion of up to 300%.The subsequent extraction of lithium-ion and repeated charge-discharge cycles will lead to the pulverization of the anode and the destruction of the solid electrolyte interface (SEI).At present, a variety of processes have been developed to solve the problem of anisotropic volume expansion, one of which is the porous siliconbased materials [2].
In the past 60 years, dozens of processes for producing porous silicon have been developed.From the perspective of raw materials, these processes can be divided into two categories: bottom-up and topdown [3].The former uses silicon-containing compounds as raw materials and often uses reduction reactions at high temperatures to prepare porous silicon.These processes are easy to scale, but often involve high-energy consumption processes, and their competitiveness will be limited in the context of sustainable development [4].In contrast, many top-down processes can be carried out under mild conditions.This kind of process starts with the Si wafer or particles and mainly forms a porous structure through etching, including electrochemical etching, stain etching, metal-assisted chemical etching (MACE), etc [5].However, these etching methods were originally designed for the field of microelectronics and are more suitable for silicon wafers.For example, anisotropic porous arrays formed by electrochemical etching are used for the preparation of photonic crystal devices [6].On the other hand, MACE is a highly controllable and adaptable top-down chemical process, which can be applied to almost all forms of silicon materials (including wafers, powders and particles).However, the use of silver nitrate and other precious metal salts involved in MACE will inevitably increase the cost of largescale production [7].Considering the scale of the lithium-ion battery industry is so huge and still growing rapidly, it seems necessary to develop an electroless etching process using low-cost materials to produce porous silicon powder products.
Chemical corrosion of silicon in an acidic fluoride solution is one of the most important processes for the formation of porous silicon.This process is mainly based on the hole injection by oxidant into the silicon surface in solution and the subsequent reaction with fluorine ion and electron losing Si to form soluble compounds [8].The oxidant not only needs to meet the thermodynamic requirement that the redox potential is higher than that of Si, but also has a great influence on the kinetics of this reaction.Compared with the most commonly used nitric acid and hydrogen peroxide, the hydroxyl radical has a higher redox potential (2.8 V), which is expected to be used in the chemical etching process of silicon.The Fenton reaction between iron-based materials and hydrogen peroxide in an acidic aqueous solution is one of the most commonly used methods to produce hydroxyl radicals [9].The redox potential of hydroxyl radical is second only to fluorine, which can rapidly degrade a variety of organic pollutants, and is widely used in environmental engineering [10].However, to our best knowledge, there have been no reports on the use of the Fenton reaction to generate hydroxyl radicals for the chemical etching of silicon.
Compared to commercial silicon powder which is commonly used in LIB anode fabrication, ferrosilicon, an alloy of silicon and iron, is more affordable and readily available.In this manuscript, we use ferrosilicon as the raw material, and the mixed solution of hydrofluoric acid and hydrogen peroxide as the etchant.The hydroxyl radicals generated by the Fenton reaction between iron and hydrogen peroxide are expected to promote the hole injection process, resulting in the preparation of hierarchical porous silicon-based materials with both mesopores and micropores.Additionally, the low cost of ferrosilicon ($1200~1500/ton) compared with the silicon powder of industrial grade (~$3000/ton) traditionally used in Li-ion battery anode preparation also renders this method potential and feasible for large-scale production of Si-based anode materials.

Experimental
5 g of the pristine ferrosilicon (FeSi) powder (Grade FeSi75Al1.5-A,500 mesh) is first added into 200 mL HCl (2 M), and the mixture is stirred under ambient conditions for 60 min in order to reduce the metal content for further reaction.After washing with deionized water to remove excess acid, the HCltreated Fe-Si powder (de-FeSi) is added into a mixed solution of HF (4.8 M) and H 2 O 2 (5.0 M) for preparation of porous structure.The chemical etching is proceeded at 10℃ with agitation in a mechanically refrigerated cold bath for 12 hours.Finally, the product (de-FeSi-P) is collected by centrifugation after washing with deionized water and anhydrous ethanol several times, and dried at 60℃ under vacuum for 24 hours.Two control samples are prepared by direct etching of commercial Si powder (500 mesh) and pristine FeSi using HF/H 2 O 2 solution, designated as Si-P and FeSi-P, respectively.
Scanning electron micrographs are collected under a ZEISS Sigma 300, with an accelerating voltage of 5 kV and InLens observation mode.The pore size distributions are evaluated by nitrogen adsorption (77 K) with a Quantachrome Autosorb-IQ3 system.The crystal structures of the prepared materials are characterized by a D8 ADVANCE X-ray Powder diffractometer (XRD).The elemental analysis is performed on an iCAP 7600 ICP-AES.
Electrochemical measurements are conducted at 25℃ using CR2025 coin cells assembled in an argon glove box.The working electrode is prepared as a mixture of 70% active substance, 20% CMC, and 10% Super P, which is loaded with a mass of approximately 1.0 mg/cm 2 .A lithium tablet was utilized as both the counter and reference electrode, and the Celgard 2400 membrane served as the separator.The electrolyte is by solving 1 M LiPF 6 in 1:1 ethylene carbonate (EC)/diethyl carbonate (DEC) with 10 wt % FEC additives.A LAND test system is utilized to evaluate the electrochemical performance of the assembled half-cell with a 0.01-2.0V cut-off voltage.A CHI660E is used to obtain electrochemical impedance spectroscopy (EIS) and cyclic voltammograms (CVs) data.

Results and discussion
Figure 1 shows our synthetic procedure for the preparation of ferrosilicon-based hierarchical porous materials.Ferrosilicon is an iron-silicon alloy made by smelting coke, steel chips and silica as raw materials, with a high iron content and other metals such as aluminium (Al), calcium (Ca) and manganese (Mn).Therefore, before etching, the pretreatment by HCl is required to reduce the metal content to obtain dealloyed ferrosilicon powder.The results of the elemental analysis show that after dealloying by HCl treatment, the iron content is reduced from 23% to 15%.For the other elements, the contents of Al, Ca and Mn are reduced from 1.1%, 0.9% and 0.6% to 0.5%, 0.2% and 0.2%, respectively.The dealloyed Fe-Si powder is then etched in a mixture of HF and H 2 O 2 .In this acidic environment, ferrous ions can be continuously supplied by the residual iron, which later reacts with H 2 O 2 to form hydroxyl radicals (OHꞏ) [11].Figure 2 shows the scanning electron microscopy images of each prepared sample.As shown by Figure 2a and Figure 2b, the surface of pristine ferrosilicon particles is fairly smooth and their size is widely ranging from 50 microns to 100 nm.After dealloying with HCl, some columnar and shallow pored are found on the particle surface, which is the result of the partial removal of iron-rich phases (Figure 2c).Upon subsequent etching with HF/ H 2 O 2 , the morphology of ferrosilicon particles has transformed into a sponge-like porous structure (Figure 2d).As shown by the BJH pore size distribution data in Figure 3, in addition to the macropores ranging from 100-300 nm found by SEM, microporous structures with a pore size ranging from 1.5-4 nm can be formed in this etching process.The morphology of the control sample prepared from pure Si particles (Si-P) is shown in Figure 2e for further investigation of the effect of OHꞏ generated by the Fenton reaction in HF/H 2 O 2 solution.Similar macroporous frameworks are observed for Si-P.However, a very different pore size distribution without any micropores is given by nitrogen adsorption as shown in Figure 3a.The microporous structures found in de-FeSi-P should therefore be attributed to the presence of iron components, namely the hole injection by OHꞏ generated by the Fenton reaction.Furthermore, another control sample FeSi-P is prepared by direct etching of pristine ferrosilicon powder in HF/H 2 O 2 solution without HCl treatment, in order to check the necessity of the dealloying process.The irregular morphology of FeSi-P (Figure 2f) shows that an excessive metal content will lead to the failure of porous structure formation.This indicates the fine-tuning of metal content during the dealloying process is necessary for obtaining optimized porous structures in the subsequent etching step, which will be studied in detail in our future research.FeSi powder, de-FeSi, de-FeSi-P and FeSi-P.Figure 3b shows the X-ray diffraction data of pristine FeSi, de-FeSi, de-FeSi-P and FeSi-P.The peaks located at 2θ values of 17.25, 37.67 and 48.97° are ascribed to the (001), ( 101) and (102) planes of crystalline FeSi 2 , respectively [12].These peaks are weakened but still observable in de-FeSi and de-FeSi-P owning to the dealloying process.On the other hand, the peaks located at 2θ values of 28.44, 47.30, 56.12, 69.13, 76.38 and 88.03° in all the samples are assigned to the (111), ( 220), (331), (400), (331) and (422) planes of silicon crystals, respectively [13].According to these XRD data, the main crystal structures are not affected by the etching process, which is highly favoured by scaled production.
In order to investigate the applicability of de-FeSi-P for lithium-ion battery anode, a CR2025 halfcell is assembled and the first four cyclic voltammetry curves at a scan rate of 0.1 mV/s from 0.01V to 1.0 V (versus Li/Li+) are shown in Figure 4a.Two anodic peaks are clearly observed at 0.35 V and 0.53 V for all four cycles, respectively, which can be assigned to the transition of Li x Si to Li during the Li extraction process [14].The increasing intensity of these two peaks indicates the gradual activation process during the cycles and the existence of crystalline silicon phase in de-FeSi-P, which is consistence ICAMIM-2023 Journal of Physics: Conference Series 2720 (2024) 012043 with XRD data in Figure 3b.The cathodic peak observed at 0.15 V during the charging process can be assigned to the formation of the amorphous Li x Si phase during the Li insertion process [15].Figure 4b shows the EIS curves of Si-P and de-FeSi-P for the comparison of the transportation properties of electrodes.The R ct of de-FeSi-P in the equivalent circuit model, corresponding to the charge-transfer resistance, is calculated as 82 Ω, much lower than that of Si-P (311 Ω).Inferred from the above experimental results, the residual FeSi 2 phase in de-FeSi-P can improve the electric performance without interfering with the insertion and extraction of Li ions, which is, therefore, a high potential candidate for LIB anode.

Conclusion
In conclusion, a low-cost porous material is successfully prepared by Fenton reaction-assisted chemical etching of partially dealloyed ferrosilicon powder in HF/H 2 O 2 .The electrochemical results show that this ferrosilicon-based hierarchical structure is capable of smooth lithium-ion insertion and extraction as a demonstration of the concept and feasibility for LIB application.Compared with the porous anode based on traditionally used commercial Si powder, this material has fine porous frameworks with pore size mainly ranging from 2 nm to 4 nm and a much lower charge-transfer resistance owing to the high conductivity of the partially retained FeSi 2 phase.Additionally, the low cost of ferrosilicon ($1200~1500/ton) compared with commercial silicon powder (~$3000/ton) traditionally used for anode preparation also renders this method attractive for large-scale production of Si-based anode materials.

Figure 1 .
Figure 1.Fabrication process of porous framework based on Fenton reaction assisted etching of dealloyed ferrosilicon.Figure2shows the scanning electron microscopy images of each prepared sample.As shown by Figure2aand Figure2b, the surface of pristine ferrosilicon particles is fairly smooth and their size is widely ranging from 50 microns to 100 nm.After dealloying with HCl, some columnar and shallow pored are found on the particle surface, which is the result of the partial removal of iron-rich phases (Figure2c).Upon subsequent etching with HF/ H 2 O 2 , the morphology of ferrosilicon particles has transformed into a sponge-like porous structure (Figure2d).As shown by the BJH pore size distribution data in Figure3, in addition to the macropores ranging from 100-300 nm found by SEM, microporous structures with a pore size ranging from 1.5-4 nm can be formed in this etching process.The morphology of the control sample prepared from pure Si particles (Si-P) is shown in Figure2efor further investigation of the effect of OHꞏ generated by the Fenton reaction in HF/H 2 O 2 solution.Similar macroporous frameworks are observed for Si-P.However, a very different pore size distribution without any micropores is given by nitrogen adsorption as shown in Figure3a.The microporous structures found in de-FeSi-P should therefore be attributed to the presence of iron components, namely the hole injection by OHꞏ generated by the Fenton reaction.Furthermore, another control sample FeSi-P is prepared by direct etching of pristine ferrosilicon powder in HF/H 2 O 2 solution without HCl treatment, in order to check the necessity of the dealloying process.The irregular morphology of FeSi-P (Figure2f) shows that an excessive metal content will lead to the failure of porous structure formation.This indicates the fine-tuning of metal content during the dealloying process is necessary for obtaining optimized porous structures in the subsequent etching step, which will be studied in detail in our future research.

Figure 3 .
Figure 3. (a) Pore size distribution curves of de-FeSi-P and Si-P, (b) XRD patterns of pristineFeSi powder, de-FeSi, de-FeSi-P and FeSi-P.Figure3bshows the X-ray diffraction data of pristine FeSi, de-FeSi, de-FeSi-P and FeSi-P.The peaks located at 2θ values of 17.25, 37.67 and 48.97° are ascribed to the (001), (101) and (102) planes of crystalline FeSi 2 , respectively[12].These peaks are weakened but still observable in de-FeSi and de-FeSi-P owning to the dealloying process.On the other hand, the peaks located at 2θ values of 28.44, 47.30, 56.12, 69.13, 76.38 and 88.03° in all the samples are assigned to the (111), (220), (331), (400), (331) and (422) planes of silicon crystals, respectively[13].According to these XRD data, the main crystal structures are not affected by the etching process, which is highly favoured by scaled production.In order to investigate the applicability of de-FeSi-P for lithium-ion battery anode, a CR2025 halfcell is assembled and the first four cyclic voltammetry curves at a scan rate of 0.1 mV/s from 0.01V to 1.0 V (versus Li/Li+) are shown in Figure4a.Two anodic peaks are clearly observed at 0.35 V and 0.53 V for all four cycles, respectively, which can be assigned to the transition of Li x Si to Li during the Li extraction process[14].The increasing intensity of these two peaks indicates the gradual activation process during the cycles and the existence of crystalline silicon phase in de-FeSi-P, which is consistence

Figure 4 .
Figure 4. (a) CV curves for de-FeSi-P at 0.1 mV/s (b) EIS curves for Si-P and de-FeSi-P electrodes.