Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2020 Jan 31;5(5):2081–2087. doi: 10.1021/acsomega.9b01665

Effects of HF on the Lithiation Behavior of the Silicon Anode in LiPF6 Organic Electrolyte Solution

Huiwen Lin †,‡, Hidenori Noguchi †,‡,*, Kohei Uosaki ‡,*
PMCID: PMC7016919  PMID: 32064368

Abstract

graphic file with name ao9b01665_0007.jpg

As one of the major impurities in the organic electrolyte, HF can react with the alkali components in the solid electrolyte interphase (SEI), such as lithium alkoxide and lithium carbonate, to form more LiF-rich SEI. Here, the effects of HF on the lithiation behavior of the single crystal Si(111) anode were studied using scanning electron microscopy, soft X-ray emission spectroscopy, and windowless energy-dispersive X-ray spectroscopy. When the Li–Si alloy is formed in 1.0 M LiPF6 in the propylene carbonate solvent, it has a layered structure that contained the first layer of crystalline Li15Si4 (c-Li15Si4) alloy pyramids, the second layer of amorphous Li13Si4 (a-Li13Si4) alloy, and a third layer of Li-diffused LixSi alloy. When the more concentrated HF is in the electrolyte solution, less amount of the c-Li15Si4 alloy is formed in the first layer. It suggests that the Si lithiation can form only amorphous LixSi alloy relative to the components in the electrolytes. The study also explains why only amorphous LixSi alloy formation was observed in some previous studies.

1. Introduction

Si is one of the most promising candidates among all the anode materials investigated for Li-based secondary batteries because of its high theoretical capacity of 3600 mA·h·g–1 when a fully alloyed Li15Si4 phase forms at room temperature. However, a large volume expansion of close to 400% prevents the application of Si anodes in Li-based secondary batteries.1−3 The solid electrolyte interphase (SEI) between the Si anode and electrolyte can be destroyed because of the expansion and shrinkage of the Si anode during a cycle.4 The physical and chemical stability of the SEI can affect battery performance in terms of the cycle life and capacity.5−8 An unstable SEI is detrimental to the cycle life of secondary batteries because the electrolyte is consumed because of the reaction with the Si anode. Several strategies, such as the use of Si nanoparticles or nanowires, have proven to be effective to reduce the volume expansion to reduce damage to the SEI.9−11 In addition, the use of electrolyte additives is also an effective strategy to form a stable SEI to retard the electrolyte decomposition.12,13

In general, there are some impurities such as H2O or HF that cannot be eliminated completely in the organic electrolyte. LiPF6 is commonly used in carbonate-based electrolytes because of its high solubility in the organic solvent. LiPF6 is thermally unstable when there is trace H2O and easily decomposes to generate POF3 and another impurity HF, which will further etch SiO2 in the Si anode.14,15 As an impurity, HF has some beneficial effects on SEI stability in the Li-based secondary batteries. It can react with the alkali components in the SEI, such as lithium alkoxide and lithium carbonate, to form a more LiF-rich SEI which can protect the Si anode.16−19 However, little effort has been devoted to study the effect of HF on the lithiation behavior of Si anodes, especially on the structures and compositions of Li–Si alloys. Our previous studies had verified the availability of using soft X-ray emission spectroscopy (SXES) and windowless energy-dispersive X-ray spectroscopy (EDS) to determine the structures and compositions of electrochemically formed Li–Si alloys. It is revealed that the Li–Si alloy has a layered structure and composition, which contains c-Li15Si4, a-Li13Si4, and Li-diffused LixSi.20,21 The determined composition consists of three different alloys and differs from other studies that only reported the c-Li15Si4 alloy22 or only reported the amorphous LixSi alloy23,24 formation. In this work, we move forward to study how HF, one of the major impurities in the electrolyte, affects the structures and compositions of the Li–Si alloy by using these techniques.

2. Results and Discussion

Figure 1a,b shows scanning electron microscopy (SEM) images of the surface and the cross section of Li–Si alloys formed in the 1.0 M LiPF6/PC electrolyte. Figure 1a shows that a large number of pyramid structures (inset figure in Figure 1a) covered the Si surface. Figure 1b shows the layered structure of the Li–Si alloy. The first layer consists of triangular pyramid structures with a height of approximately 0.7 μm. Underneath the first layer, there are the second and third layers distinguished by the SEM image contrast, respectively.

Figure 1.

Figure 1

Open in a new tab

SEM images of (a) surface and (b) cross section of Li–Si alloys formed in electrolytes of 1.0 M LiPF6 dissolved in PC containing 0.05 mM HF. In (b), the layered structures of the first, second, and third layers and Si substrate are represented by different colored arrows and numbers. The arrow lengths are taken from the results of the windowless EDS in Figure 4. The compositions are shown according to the SXES results in (c). (c) SXE spectra for the first, second, third, and fourth layers of the Li–Si alloy. The peak energies of 53.6 eV in Li Kα and 91.6 eV in Si L2,3 relative to the a-Li13Si4 alloy are indicated. The peak energies of 53.3 eV in Li Kα and 91.8 eV in Si L2,3 relative to the c-Li15Si4 alloy are also indicated.

The compositions of the Li–Si alloys were determined using SXES. Figure 1c shows SXE spectra for the first, second, and third layers, in addition to those for Li metal and c-Si as references22,23 in the photon energy region of 50–105 eV, which covers Li Kα and Si L2,3 X-ray emissions. Combined with density of states calculations,25 the peak energy of 53.3 eV in Li Kα and 91.8 eV in Si L2,3 has been indicated to be relative to the c-Li15Si4 alloy, and a peak energy of 53.6 eV in Li Kα and 91.6 eV in Si L2,3 is relative to the a-Li13Si4 alloy.20 The peak energies were used to determine the compositions of the Li–Si alloys. The Li Kα and Si L2,3 emission peak energies of the first layer in the Li–Si alloys formed in propylene carbonate (PC) were 53.3 and 91.8 eV, respectively, which determined that the first layer was the c-Li15Si4 alloy. The Li Kα and Si L2,3 emission peak energies of the second layer in the Li–Si alloys formed in PC were 53.6 and 91.6 eV, respectively, which determined that the second layer was the a-Li13Si4 alloy. There was little Li Kα emission intensity for the third layers because the Li concentration in the third layer was very low.21 Therefore, the determined composition of the Li–Si alloy formed in PC were consistent with that reported by Aoki et al.,20 which determined that the first layer was a c-Li15Si4 alloy and the second layer was a-Li13Si4 alloy, although they used a different solvent [v(ethylene carbonate, EC)/v(dimethyl carbonate, DMC) = 1:1]. The composition of the third layer was difficult to determine from the SXE spectra. It was previously determined to be a Li-diffused LixSi alloy layer using windowless EDS, although the solvent is DMC in that work.21

As mentioned above, HF and water are two of the major impurities that cannot be eliminated completely in the LiPF6-containing organic electrolyte. The HF concentration in the 1.0 M LiPF6/PC electrolyte was determined to be less than 0.05 mM using IC. The water content in the electrolytes were determined to be approximately 0.2 mM. In the following section, HF aqueous solution was added in the electrolyte of 1.0 M LiPF6/PC to study the effects of HF on the lithiation behavior of the Si anode. Figure 2a–d shows SEM images of the surface of Li–Si alloys formed in different electrolytes with more concentrated HF. The HF concentration are shown along with the water concentration. As is observed, the pyramid structures are present in all of the Li–Si alloys, and their size is dependent on the added amount of HF aqueous solution in the electrolytes. Because 46% HF aqueous solution was used, water was also added when we tried to increase the HF concentration. To make sure water had no effects on the pyramid formation, only ultrapure water was added into the electrolyte and the Li–Si alloy formed therein is shown in Figure 2e. Here, the water concentration of 2.2 mM for Figure 2e is the same as that for Figure 2d, but the HF concentration is only 0.05 mM. Comparing Figure 2a,e, it was found that only 2 mM water addition has negligible effect on the structure of the Li–Si alloy and does not change the pyramid size. The sizes of the pyramid structures that are dependent on the HF concentrations in the electrolytes are shown in Figure 2f. When the more concentrated HF is in the electrolyte solution, the pyramid size of the Li–Si alloys decreased. The effect of HF concentration on the lithiation behavior of Si is also reflected in the current–time profiles, as shown in Figure 2g. It is observed that all the samples have a maximum lithiation current. The maximum lithiation current is corresponding to the time when the pyramids grow to overlap each other. The maximum lithiation current and its corresponding time decreased when the HF concentration increased as indicated by the arrow. Because when the HF concentration increased, more pyramids (the number of pyramids increases in the same area as shown in Figure 2) are formed to overlap each other earlier. Therefore, an earlier overlap of the pyramids leads to the decrease of the pyramid size when the HF concentration increases.

Figure 2.

Figure 2

Open in a new tab

(a–d) SEM images of the surfaces of Li–Si alloys formed in 0.05/0.5/1/2 mM HF containing electrolytes. The water content is also shown. (a) Enlarged region of Figure 1a. (e) SEM images of the surfaces of Li–Si alloys formed in the electrolyte that contains 2.2 mM water and 0.05 mM HF. (f) Change in the size (solid dots with an error bar) of the pyramid structures with the HF concentration for the Li–Si alloys as shown in (a–d). At least 100 pyramids were measured. (g) Current–time profile of the Si lithiation in the increased HF concentration.

The effects of HF can be better determined by observation of its effects on the layered structures of the Li–Si alloys. Figure 3a–d shows SEM images of the cross section of the Li–Si alloys. Layered structures were observed for all of the Li–Si alloys. The Li–Si alloy formed in the 1.0 M LiPF6/PC electrolyte without aqueous HF addition was determined to consist of the first layer of crystalline Li15Si4, the second amorphous Li13Si4 layer, and a third Li-diffused LixSi layer. The layer thicknesses of the first and second layers in different Li–Si alloys were dependent on the electrolytes with increased HF concentrations.

Figure 3.

Figure 3

Open in a new tab

(a–d) SEM images of the cross sections of Li–Si alloys formed in the electrolytes that contain 0.05/0.5/1/2 mM HF. (a) Enlarged region of Figure 1b. (e) Li Kα and (f) Si L2,3 emission spectra for the first layer in the four Li–Si alloys. (g) Li Kα and (h) Si L2,3 emission spectra for the second layer in the four Li–Si alloys. The peak energy positions are represented by dotted lines; the 53.3 and 53.6 eV peaks in the Li Kα emission spectra and the 91.6 and 91.8 eV peaks in the Si L2,3 emission spectra.

The effects of HF on the compositions of the Li–Si alloys were determined using SXES. Figure 3e,f shows SXE spectra for the first layers in the Li–Si alloys formed in the electrolytes with increased HF concentrations. When the more concentrated HF is in the electrolyte solution, the Li Kα emission peak of the Li–Si alloys changed from 53.3 to 53.6 eV (Figure 3e). The peaks at 53.3 and 53.6 eV were the peak energies of the Li Kα emission spectra for c-Li15Si4 and a-Li13Si4, respectively, according to previous analyses.20 The change in the peak energy from 53.3 to 53.6 eV indicated that the amount of c-Li15Si4 in the first layer was decreased, and instead, the amount of a-Li13Si4 increased. The Si L2,3 emission peak of the Li–Si alloys changed from 91.8 to 91.6 eV (Figure 3f). The peak energies of the Si L2,3 emission spectra of c-Li15Si4 and a-Li13Si4 were 91.8 and 91.6 eV, respectively; therefore, the peak shift from 91.8 to 91.6 eV indicated that the first layer preferred to form a-Li13Si4 instead of c-Li15Si4. This result is consistent with that from the observed shift of the Li Kα emission peak, that is, the amount of c-Li15Si4 in the first layer decreased when the HF concentration in the electrolyte increased. The compositions of the Li–Si alloys were thus altered by HF addition.

Figure 3g,h shows SXE spectra for the second layers in the Li–Si alloys. Although the electrolytes had more concentrated HF, the Li Kα and Si L2,3 emission peaks of the Li–Si alloys had the same energy positions of 53.6 and 91.6 eV, respectively. Both the peak positions correspond to the Li Kα and Si L2,3 emission peaks of a-Li13Si4. The results indicated that the composition in the second layer of all four different Li–Si alloys was a-Li13Si4 and was independent of the HF concentration.

Windowless EDS can be applied to determine the Li distribution and the layer thickness in Li–Si alloys. For the Li–Si alloy formed in the 1.0 M LiPF6/PC electrolyte, the Li distribution profile had been determined by windowless EDS and was consistent with the layered structure determined by SXES.21 The SXE spectral analysis suggested that the compositions of the Li–Si alloys were different when they were electrochemically formed in different HF concentrations. As shown in Figure 4a, the Si L2,3 emission profiles show a distinguishable intensity change along the layered structure. Although there is some contrast difference with the third layer, we determine it as a layer because it has almost no Li Kα emission intensity and its exact composition cannot be accurately determined. Windowless EDS was used here to further reveal the extent of the different compositions, that is, c-Li15Si4, a-Li13Si4, and Li-diffused LixSi alloy. Figure 4b,c shows the changes in the intensities of the Li Kα and Si Kα emissions along the line scan from the pyramid top (the first layer) to the Si substrate for the Li–Si alloys, respectively. The changes in the Li Kα intensities were similar for all four Li–Si alloys in Figure 4b. The Li Kα intensity decreased from the first layer to the Si substrate, which is consistent with the Li–Si alloy-layered structure that consists of the first c-Li15Si4 layer, the second a-Li13Si4 layer, and a third lowly lithiated LixSi layer over the Si substrate. The boundaries between the first and second layers and between the second and third layers can be determined by monitoring the change in the intensity of the Li Kα peak. It can be easily observed whether the layer boundaries (dotted arrows) in the Li–Si alloys moved to lesser depths when the more concentrated HF is in the electrolyte solution. In Figure 4c, the Si Kα intensity increased from the first layer to the Si substrate. The boundaries between the second and third layers could be determined by monitoring the change in the intensity of the Si Kα peak. It was also observed that the layer boundaries moved to lesser depths when the more concentrated HF is in the electrolyte solution, which is consistent with the observed change in the intensity of the Li Kα peak. However, changes in the intensities of the Li Kα and Si Kα peaks could not be used to determine the thickness of each layer in the Li–Si alloys because the boundaries between the third layer and Si substrate were unclear.

Figure 4.

Figure 4

Open in a new tab

(a) Si L2,3 emission profiles in windowless EDS along the cross section of the Li–Si alloy that was formed in the 1.0 M LiPF6/PC electrolyte with 0.05 mM HF. SEM image is part of Figure 3a that is also used for SXE measurement. Windowless EDS (b) Li Kα, (c) Si Kα, and (d) Si L2,3 emission profiles along the cross section from the pyramid top to the Si substrate for the four Li–Si alloys. The HF concentrations in the electrolytes are shown. The boundary between the first, second, and third layers, and the Si substrate can be determined by monitoring the intensity change. In (d), the cyan arrows represent the boundary between the first and second layers, the dark yellow arrows represent the boundary between the second and third layers, and the dark green arrows represent the boundary between the third and fourth layer. (e) Layer thicknesses of the first, second, and third layers for the four Li–Si alloys determined from (d).

As reported previously, the thickness of each layer can be precisely determined by monitoring the change in the Si L2,3 intensity.21Figure 4d shows the Si L2,3 intensity changes for the four Li–Si alloys. The boundaries between the first, second, and third layers, and the Si substrate can be easily determined by monitoring the change in intensity, as shown by the arrows. The thickness of each layer for all the four Li–Si alloys can then be easily determined, and the results are shown in Figure 4e. When the electrolytes had more concentrated HF from 0 to 2 mM, the first layer’s thickness (the c-Li15Si4 pyramid height) in the Li–Si alloy decreased from approximately 0.7 to 0.1 μm, and the second layer’s thickness increased from approximately 1.0 to 1.3 μm. The Li distribution in the Li–Si alloys was thus changed when the more concentrated HF is in the electrolyte solution. The thicknesses of the third layers were almost the same for all the four Li–Si alloys.

The thickness t can be used to estimate the amount of c-Li15Si4, a-Li13Si4, and Li-diffused LixSi alloy in the Li–Si alloys according to the equation of Q = x·F·V·ρ/(M·A), where Q and A are the charge density and the electrode area; x, V, ρ, and M are the number of Li atoms, the volume, the mass density, and the molar mass of LixSiy alloys, respectively. For c-Li15Si4, the volume V is equal to Inline graphic because it is a pyramid. For a-Li13Si4, the volume V is A·t. The lithiation charge density was 1000 mC·cm–2 for each Li–Si alloy. The lithiation charge was due to Li+ ion reduction to alloy Li; therefore, the amount of Li in c-Li15Si4 and a-Li13Si4 can be evaluated based on their layer thicknesses to reveal how much charge was used to form c-Li15Si4 and a-Li13Si4. The Li–Si alloy formed in the 1.0 M LiPF6/PC electrolyte without HF addition had approximately 0.7 μm thick c-Li15Si4 (pyramid) and 1.0 μm thick a-Li13Si4 layers. Therefore, the charges used to form c-Li15Si4 and a-Li13Si4 were approximately 180 and 760 mC·cm–2, respectively, in 1000 mC·cm–2 charge, that is, ca. 18 and 76% of the charge were used to form c-Li15Si4 and a-Li13Si4, respectively. The remaining 6% charge was used for the formation of the third layer and electrolyte decomposition. As a comparison, the Li–Si alloy formed in 1.0 M LiPF6/PC with 2 mM HF addition had approximately 0.1 μm thick c-Li15Si4 and 1.3 μm thick a-Li13Si4 layers. The charges used to form c-Li15Si4 and a-Li13Si4 were approximately 30 and 900 mC·cm–2, respectively, that is, ca. 3 and 90% charges were used to form c-Li15Si4 and a-Li13Si4, respectively. The amount of c-Li15Si4 in the Li–Si alloys decreased from 18 to 3% and the amount of a-Li13Si4 in the Li–Si alloys increased from 76 to 90% when the electrolyte used had 2 mM HF addition, as shown in Figure 4e. Formation of the c-Li15Si4 alloy was generally considered to be detrimental to battery performance;26,27 therefore, less formation of c-Li15Si4 in the Li–Si alloy by HF will be expected to be good for battery performance, although HF is generally considered to be harmful in battery and needs to be eliminated. However, a low but suitable amount of HF may be used for the Si-based secondary battery.

It is now known that HF can reduce the size of the c-Li15Si4 pyramid structures, that is, increase the number density of the pyramid nuclei. The number density of the pyramid nuclei can be affected by surface electrochemical conditions, such as the surface electrochemical potential or surface ion concentration. Here, HF addition may change the surface electrochemical conditions, which then changes the number density of the pyramid nuclei. It has been reported that HF can be adsorbed on a Si surface to form F–Si bonds.28−30 The adsorption of F species on Si(111) was confirmed by X-ray photoelectron spectroscopy (XPS) measurements, as shown in Figure 5a. The photoelectron emission intensities from F 1s increased when the more concentrated HF is in the electrolyte solution, which indicated that the amount of F on the Si(111) surface increased. The increase in the amount of the F element was due to both enhanced LiF (686.5 eV) deposition through Li+ and F– ions, and to F atom adsorption on the Si(111) surface to generate F–Si bonds (688.5 eV) as shown in Figure 5b. An intensity reverse of LiF and F–Si in the 0.5 mM HF sample might be due to the difference in the formation of the two species. LiF forms in the solution and deposits onto the Si surface within an uncontrolled amount, but F–Si forms on the Si surface and its amount is expected to be relative to HF concentration. Therefore, there is a possibility that LiF deposit more under uncontrolled amount and the intensity reverse appear. The adsorption of the F atom on H-terminated Si(111) can be through the reaction of HF (aq) + SiH (a) = SiF (a) + H2 (g) or HF (aq) + Si–Si (a) = SiF (a) + SiH (a) in which HF attacks the Si–Si bond as shown in Figure 6. These reactions suggest adsorption of the F atom to replace H and a strong SiF chemical bond formation on Si(111).

Figure 5.

Figure 5

Open in a new tab

(a) XPS F 1s spectra for Si(111) after contact with electrolytes that had increased HF concentrations. Peaks at 688.5 eV are representative of F–Si species, while those at 686.5 eV are representative of LiF species. (b) XPS peak areas at 688.5 eV for F–Si on Si(111) after contact with electrolytes that had more concentrated HF. The peak area is normalized to Si 2p intensity.

Figure 6.

Figure 6

Open in a new tab

Schematic to illustrate that the Li–Si alloys have smaller pyramid structures (lesser amount of c-Li15Si4) when the electrolyte is 1.0 M LiPF6/PC with more concentrated HF.

Fluorine adsorption can increase the formation of the Li–Si alloy nuclei because it increases the concentration of surface Li+ ions, which is a function of surface potential according to the equation31

2.

where Cs is the concentration of surface Li+ ions, C0 is the bulk Li+ ion concentration, Ø is the surface potential, and F, R, and T are the Faraday constant, gas constant, and room temperature, respectively. The surface potential Ø becomes more negative by F adsorption, as schematically shown in Figure 6; therefore, the concentration of surface Li+ ions increases. As a result, the number density of the Li–Si alloy nuclei on the Si surface increases, which results in the formation of smaller pyramid structures. This explains why the Li–Si alloys have smaller pyramid structures (lesser amount of c-Li15Si4) when the electrolyte is 1.0 M LiPF6/PC with a higher HF concentration. In addition to F adsorption to increase the nuclei number density, the deposition of LiF on the Si surface might play a role that retards the growth of the c-Li15Si4 pyramid, although some further verification will be conducted.

3. Conclusions

The Li–Si alloy prefers to form the a-Li13Si4 alloy instead of c-Li15Si4 when the more concentrated HF is in the electrolyte solution. Through using a suitable electrolyte such as using the fluoroethylene carbonate as the co-solvent which can decompose to generate HF, the Si lithiation can form only amorphous LixSi alloy. The study explains why only amorphous LixSi alloy formation was observed in some previous studies. On the other hand, less c-Li15Si4 formation in Li–Si alloys by HF is possible for better battery performance because the crystalline alloy phase is generally considered to be detrimental to battery performance.

4. Experimental Section

4.1. Materials

H2SO4 (96–98%), H2O2 (30–35%), HCl (35%), and HF (46%) aqueous solutions and anhydrous PC solvent (>98%, battery grade) were purchased from Wako Pure Chemicals. The PC solvent was dried at least 3 days by a molecular sieve which was pre-dried in a vacuum at 200 °C for 12 h. Every time before the PC solvent was used to prepare the fresh electrolyte, PC will be filtered using a 100 nm nylon filter. NH4F aqueous solution (40%) (semiconductor grade) was purchased from Morita Chemical Industry. LiPF6 salt (99.9%) and Li wire (99%) were purchased from Kishida Chemical. LiPF6 was dried under a vacuum at room temperature for 6 h at first and then increase the temperature to 105 °C and kept for 6 h. The water vapor was trapped by a cold trap using liquid nitrogen. Si(111) single-crystal wafers (thickness: 500 μm, resistivity: 1–10 Ω·cm–1, n-type: phosphorous-doped) were purchased from a Shin-Etsu Semiconductor. Ultrapure water from a Milli-Q purification system and 99.999% Ar gas were used throughout the experiments.

4.2. Electrochemically Formed Li–Si Alloys under Potentiostatic Lithiation

The Si(111) wafer was cut into 1.0 × 1.0 cm2 pieces. Hydrogen-terminated Si(111) was prepared through chemical etching, as reported previously.32 Li wire was used as reference and counter electrodes. The electrolyte used was 1.0 M LiPF6 dissolved in the PC solvent. To study the effect of HF, electrolytes of 1.0 M LiPF6 in PC were prepared with the addition of 0, 0.5, 1.0, and 2.0 mM HF aqueous solution. For example, for the addition of 1 mM HF aqueous solution, 1.5 μL of 46% HF aqueous solution was added to 40 mL of 1.0 M LiPF6 in PC. Because HF aqueous solution was used, water was also added simultaneously. For example, 2.2 mM H2O was included when 2 mM HF aqueous solution was added. Here, the effect of H2O on the structures and compositions of the Li–Si alloy was examined separately by the addition of 2.2 mM pure H2O to the 1.0 M LiPF6/PC electrolyte. Li–Si alloys were formed under potentiostatic lithiation at 10 mV versus Li/Li+ in these electrolytes to determine the effects of HF. For each Li–Si alloy, the lithiation charge density was 1000 mC·cm–2. After electrochemical formation, the Li–Si alloys were washed with the DMC solvent and broken in half along the Si ⟨110⟩ direction with a diamond pen to expose the cross section. The samples were then transferred using a Ar-filled transfer vessel without air exposure for SXES and windowless EDS measurements.

4.3. SXES and Windowless EDS Measurements

SXES and windowless EDS measurements were conducted using a SXES system (JEOL, SS-9400SXES) and a windowless silicon drift detector (SDD) system (Oxford Instruments, X-Max Extreme), respectively. Both systems are installed in a scanning electron microscope (JEOL, JSM-7800F) under an ambient pressure of less than 10–4 Pa and using an accelerating voltage of 5.0 kV.

The incident electron spot size for SXES was approximately 1.0 μm and a varied line-space grating (JEOL, JS50XL) was applied with a resolution of approximately 0.15 eV. The probe current and acquisition time were 50 nA and 200 s, respectively. The EDS line scan method was applied by performing point detection at intervals of 100 nm along the Li–Si alloy layer. Each point accumulation time was set to 15 s because the accumulated intensity with this time was appropriate for deconvolution analysis, and the specimen was not destroyed by the electron beam within 15 s.

4.4. HF Concentration Measurements

The HF concentration in the electrolytes was measured using an ion chromatography (IC) system (Dionex, ICS-2100). The electrolytes were diluted 100 times for the purpose of accurate determination.

4.5. X-ray Photoelectron Spectroscopy

The surface species on the Si(111) wafers were investigated using a 1 × 10–9 Torr base pressure XPS spectrometer (ULVAC-PHI, VersaProbe II) equipped with an Al Kα source operated at 150 W. High-resolution scans were observed in constant energy mode with a pass energy of 46.95 eV. The size of the analysis area on all specimens was set at 220 μm × 220 μm. The Si wafer specimens for XPS measurements were prepared by immersion in the electrolytes with increased HF concentration and then washing with DMC to remove the electrolyte.

Acknowledgments

The present work was financially supported by the Japanese Ministry of Education, Culture, Sports, Science and Technology (MEXT) Program for Integrated Materials Development. The authors thank Ms. Makiko Oshida and Dr. Ayumi Narita for assistance with the SEM, SXES, windowless EDS, and XPS measurements.

The authors declare no competing financial interest.

References

  1. Kasavajjula U.; Wang C.; Appleby A. J. Nano- and bulk-silicon-based insertion anodes for lithium-ion secondary cells. J. Power Sources 2007, 163, 1003–1039. 10.1016/j.jpowsour.2006.09.084. [DOI] [Google Scholar]
  2. Su X.; Wu Q. L.; Li J. C.; Xiao X. C.; Lott A.; Lu W. Q.; Sheldon B. W.; Wu J. Silicon-Based Nanomaterials for Lithium-Ion Batteries: A Review. Adv. Energy Mater. 2014, 4, 1300882. 10.1002/aenm.201300882. [DOI] [Google Scholar]
  3. Chan C. K.; Peng H.; Liu G.; McIlwrath K.; Zhang X. F.; Huggins R. A.; Cui Y. High-performance lithium battery anodes using silicon nanowires. Nat. Nanotechnol. 2008, 3, 31–35. 10.1038/nnano.2007.411. [DOI] [PubMed] [Google Scholar]
  4. Lee Y. M.; Lee J. Y.; Shim H.-T.; Lee J. K.; Park J.-K. SEI layer formation on amorphous si thin electrode during precycling. J. Electrochem. Soc. 2007, 154, A515–A519. 10.1149/1.2719644. [DOI] [Google Scholar]
  5. Ogata K.; Salager E.; Kerr C. J.; Fraser A. E.; Ducati C.; Morris A. J.; Hofmann S.; Grey C. P. Revealing lithium–silicide phase transformations in nano-structured silicon-based lithium ion batteries via in situ NMR spectroscopy. Nat. Commun. 2014, 5, 3217. 10.1038/ncomms4217. [DOI] [PubMed] [Google Scholar]
  6. Gregorczyk K. E.; Liu Y.; Sullivan J. P.; Rubloff G. W. In Situ Transmission Electron Microscopy Study of Electrochemical Lithiation and Delithiation Cycling of the Conversion Anode RuO2. ACS Nano 2013, 7, 6354–6360. 10.1021/nn402451s. [DOI] [PubMed] [Google Scholar]
  7. Pharr M.; Zhao K.; Wang X.; Suo Z.; Vlassak J. J. Kinetics of Initial Lithiation of Crystalline Silicon Electrodes of Lithium-Ion Batteries. Nano Lett. 2012, 12, 5039–5047. 10.1021/nl302841y. [DOI] [PubMed] [Google Scholar]
  8. McDowell M. T.; Ryu I.; Lee S. W.; Wang C.; Nix W. D.; Cui Y. Studying the Kinetics of Crystalline Silicon Nanoparticle Lithiation with In Situ Transmission Electron Microscopy. Adv. Mater. 2012, 24, 6034–6041. 10.1002/adma.201202744. [DOI] [PubMed] [Google Scholar]
  9. Szczech J. R.; Jin S. Nanostructured silicon for high capacity lithium battery anodes. Energy Environ. Sci. 2011, 4, 56–72. 10.1039/c0ee00281j. [DOI] [Google Scholar]
  10. Liu X. H.; Zhong L.; Huang S.; Mao S. X.; Zhu T.; Huang J. Y. Size-Dependent Fracture of Silicon Nanoparticles During Lithiation. ACS Nano 2012, 6, 1522–1531. 10.1021/nn204476h. [DOI] [PubMed] [Google Scholar]
  11. Teki R.; Datta M. K.; Krishnan R.; Parker T. C.; Lu T.-M.; Kumta P. N.; Koratkar N. Nanostructured Silicon Anodes for Lithium Ion Rechargeable Batteries. Small 2009, 5, 2236–2242. 10.1002/smll.200900382. [DOI] [PubMed] [Google Scholar]
  12. Lin Y.-M.; Klavetter K. C.; Abel P. R.; Davy N. C.; Snider J. L.; Heller A.; Mullins C. B. High performance silicon nanoparticle anode in fluoroethylene carbonate-based electrolyte for Li-ion batteries. Chem. Commun. 2012, 48, 7268–7270. 10.1039/c2cc31712e. [DOI] [PubMed] [Google Scholar]
  13. Etacheri V.; Haik O.; Goffer Y.; Roberts G. A.; Stefan I. C.; Fasching R.; Aurbach D. Effect of Fluoroethylene Carbonate (FEC) on the Performance and Surface Chemistry of Si-Nanowire Li-Ion Battery Anodes. Langmuir 2012, 28, 965–976. 10.1021/la203712s. [DOI] [PubMed] [Google Scholar]
  14. Yang H.; Zhuang G. V.; Ross P. N. Thermal stability of LiPF6 salt and Li-ion battery electrolytes containing LiPF6. J. Power Sources 2006, 161, 573–579. 10.1016/j.jpowsour.2006.03.058. [DOI] [Google Scholar]
  15. Lux S. F.; Lucas I. T.; Pollak E.; Passerini S.; Winter M.; Kostecki R. The mechanism of HF formation in LiPF6 based organic carbonate electrolytes. Electrochem. Commun. 2012, 14, 47–50. 10.1016/j.elecom.2011.10.026. [DOI] [Google Scholar]
  16. Zhang S. S. A review on electrolyte additives for lithium-ion batteries. J. Power Sources 2006, 162, 1379–1394. 10.1016/j.jpowsour.2006.07.074. [DOI] [Google Scholar]
  17. Kanamura K.; Shiraishi S.; Takehara Z. Electrochemical deposition of very smooth lithium using nonaqueous electrolytes containing HF. J. Electrochem. Soc. 1996, 143, 2187–2197. 10.1149/1.1836979. [DOI] [Google Scholar]
  18. Kanamura K.; Shiraishi S.; Takehara Z. i. Electrochemical Deposition of Uniform Lithium on an Ni Substrate in a Nonaqueous Electrolyte. J. Electrochem. Soc. 1994, 141, L108–L110. 10.1149/1.2055169. [DOI] [Google Scholar]
  19. Tomita K.; Noguchi H.; Uosaki K. Effect of Water and HF on the Distribution of Discharge Products at Li-O2 Battery Cathode. ACS Appl. Energy Mater. 2018, 1, 3434–3442. 10.1021/acsaem.8b00584. [DOI] [Google Scholar]
  20. Aoki N.; Omachi A.; Uosaki K.; Kondo T. Structural Study of Electrochemically Lithiated Si(111) by using Soft X-ray Emission Spectroscopy Combined with Scanning Electron Microscopy and through X-ray Diffraction Measurements. ChemElectroChem 2016, 3, 959–965. 10.1002/celc.201600030. [DOI] [Google Scholar]
  21. Lin H.; Noguchi H.; Uosaki K. Application of windowless energy dispersive spectroscopy to determine Li distribution in Li-Si alloys. Appl. Phys. Lett. 2018, 112, 073903. 10.1063/1.5016211. [DOI] [Google Scholar]
  22. Obrovac M. N.; Christensen L. Structural Changes in Silicon Anodes during Lithium Insertion/Extraction. Electrochem. Solid-State Lett. 2004, 7, A93–A96. 10.1149/1.1652421. [DOI] [Google Scholar]
  23. Liu X. H.; Wang J. W.; Huang S.; Fan F.; Huang X.; Liu Y.; Krylyuk S.; Yoo J.; Dayeh S. A.; Davydov A. V.; Mao S. X.; Picraux S. T.; Zhang S.; Li J.; Zhu T.; Huang J. Y. In situ atomic-scale imaging of electrochemical lithiation in silicon. Nat. Nanotechnol. 2012, 7, 749–756. 10.1038/nnano.2012.170. [DOI] [PubMed] [Google Scholar]
  24. Chan C. K.; Ruffo R.; Hong S. S.; Huggins R. A.; Cui Y. Structural and electrochemical study of the reaction of lithium with silicon nanowires. J. Power Sources 2009, 189, 34–39. 10.1016/j.jpowsour.2008.12.047. [DOI] [Google Scholar]
  25. Lyalin A.; Kuznetsov V. G.; Nakayama A.; Abarenkov I. V.; Tupitsyn I. I.; Gabis I. E.; Uosaki K.; Taketsugu T. Lithiation Products of a Silicon Anode Based on Soft X-ray Emission Spectroscopy: A Theoretical Study. J. Phys. Chem. C 2018, 122, 11096–11108. 10.1021/acs.jpcc.8b00489. [DOI] [Google Scholar]
  26. Beaulieu L. Y.; Hatchard T. D.; Bonakdarpour A.; Fleischauer M. D.; Dahn J. R. Reaction of Li with Alloy Thin Films Studied by In Situ AFM. J. Electrochem. Soc. 2003, 150, A1457–A1464. 10.1149/1.1613668. [DOI] [Google Scholar]
  27. Misra S.; Liu N.; Nelson J.; Hong S. S.; Cui Y.; Toney M. F. In SituX-ray Diffraction Studies of (De)lithiation Mechanism in Silicon Nanowire Anodes. ACS Nano 2012, 6, 5465–5473. 10.1021/nn301339g. [DOI] [PubMed] [Google Scholar]
  28. Craig B. I.; Smith P. V. The chemisorption of HF and F2 on the Si(100) surface. Surf. Sci. 1990, 239, 36–41. 10.1016/0039-6028(90)90616-g. [DOI] [Google Scholar]
  29. Michalak D. J.; Amy S. R.; Aureau D.; Dai M.; Estève A.; Chabal Y. J. Nanopatterning Si(111) surfaces as a selective surface-chemistry route. Nat. Mater. 2010, 9, 266–271. 10.1038/nmat2611. [DOI] [PubMed] [Google Scholar]
  30. Saito Y. Adsorption of Anhydrous Hydrogen Fluoride onto Silicon and Native Oxide by XPS. Surface Science Spectra 1999, 6, 228–236. 10.1116/1.1247921. [DOI] [Google Scholar]
  31. Bard A. J.; Faulkner L. R.. Electrochemical Methods: Fundamentals and Applications; Wiley, 2000. [Google Scholar]
  32. Ye S.; Ichihara T.; Uosaki K. Attenuated total reflection Fourier transform infrared spectroscopy study of the adsorption of organic contaminants on a hydrogen-terminated Si(111) surface in air. Appl. Phys. Lett. 1999, 75, 1562–1564. 10.1063/1.124755. [DOI] [Google Scholar]

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES