Systematic Investigation of Binders for Silicon Anodes: Interactions of

May 2, 2016 - The cycle life of Si was greatly improved by using poly(acrylic acid) (PAA),(6) alginate extracted from brown algae,(7) and cross-linked...
32 downloads 10 Views 7MB Size
Research Article www.acsami.org

Systematic Investigation of Binders for Silicon Anodes: Interactions of Binder with Silicon Particles and Electrolytes and Effects of Binders on Solid Electrolyte Interphase Formation Cao Cuong Nguyen,† Taeho Yoon,† Daniel M. Seo,† Pradeep Guduru,‡ and Brett L. Lucht*,† †

Department of Chemistry, University of Rhode Island, Kingston, Rhode Island 02881, United States Division of Engineering, Brown University, 182 Hope Street, Providence, Rhode Island 02912, United States



ABSTRACT: The effects of different binders, polyvinylidene difluoride (PVdF), poly(acrylic acid) (PAA), sodium carboxymethyl cellulose (CMC), and cross-linked PAA−CMC (c− PAA−CMC), on the cycling performance and solid electrolyte interphase (SEI) formation on silicon nanoparticle electrodes have been investigated. Electrodes composed of Si−PAA, Si− CMC, and Si−PAA−CMC exhibit a specific capacity ≥3000 mAh/g after 20 cycles while Si−PVdF electrodes have a rapid capacity fade to 1000 mAh/g after just 10 cycles. Infrared spectroscopy (IR) and X-ray photoelectron spectroscopy (XPS) reveal that PAA and CMC react with the surface of the Si nanoparticles during electrode fabrication. The fresh Si−CMC electrode has a thicker surface coating of SiOx than Si−PAA and Si−PAA−CMC electrodes, due to the formation of thicker SiOx during electrode preparation, which leads to lower cyclability. The carboxylic acid functional groups of the PAA binder are reactive toward the electrolyte, causing the decomposition of LiPF6 and dissolution of SiOx during the electrode wetting process. The PAA and CMC binder surface films are then electrochemically reduced during the first cycle to form a protective layer on Si. This layer effectively suppresses the decomposition of carbonate solvents during cycling resulting in a thin SEI. On the contrary, the Si−PVDF electrode has poor cycling performance and continuous reduction of carbonate solvents is observed resulting in the generation of a thicker SEI. Interestingly, the Lewis basic −CO2Na of CMC was found to scavenge HF in electrolyte. KEYWORDS: silicon, binders, XPS, IR, SEI, formation retained capacity for only 8 cycles.4 The authors suggest that the addition of SBR helps to enhance the elasticity of the laminate since CMC is extremely brittle. Li et.al,5 however, showed that the Si with only CMC binder has better performance than either the SBR and CMC mixture or conventional PVdF binder. The cycle life of Si was greatly improved by using poly(acrylic acid) (PAA),6 alginate extracted from brown algae,7 and cross-linked PAA with CMC.8 In general, these binders are very stiff and contain a large quantity of −OH and −COOH functional groups. The formation of a covalent bond and/or strong interaction of hydrogen bonds between silicon particles and binders are believed to be the reason for the improvement.10−14 Recently, Si electrodes prepared with hyperbranched polysaccharides showed excellent adhesion due to strong ion−dipole interactions between the binders and silicon nanoparticles as well as the binders and copper current collectors. As a result, the electrodes showed significant improvement in capacity retention compared to CMC and alginate.15

1. INTRODUCTION Silicon is one of the most promising anode materials because of its high theoretical capacity of 3579 mAh/g (Li15Si4) which is ten times higher than graphite, the most common anode currently used in lithium ion batteries.1 Si and Si-based anodes, however, suffer from massive volume change during cycling which leads to the pulverization of active particles and loss of electrical contact between electrode components.2,3 Furthermore, the huge volume changes cause the continuous breakdown and subsequent reformation of the protective solid electrolyte interphase (SEI). As a result, the capacity of Si fades rapidly. The cycling performance of Si anodes has been found to be dependent on the structure of the polymer binders used in the composite electrode.4−8 Poly(vinylidene fluoride) (PVdF) has been widely used as binder in lithium ion batteries due to its good electrochemical and thermal stability and good adhesion to the electrode components.9 The performance of Si with PVdF, however, has been reported to be poor.5,9 Liu et al.4 and Buqa et al.9 found that the capacity retention of Si was significantly improved by using styrene butadiene rubber (SBR) and sodium carboxymethyl cellulose (CMC) as binders. The Si anode with SBR and CMC showed good capacity retention (600 mAh/g after 50 cycles) while the electrode with PVdF © 2016 American Chemical Society

Received: March 18, 2016 Accepted: May 2, 2016 Published: May 2, 2016 12211

DOI: 10.1021/acsami.6b03357 ACS Appl. Mater. Interfaces 2016, 8, 12211−12220

Research Article

ACS Applied Materials & Interfaces

A/g) for an additional 20 cycles using an Arbin BT2000 battery cycler at 25 °C.The rate was calculated on the basis of the theoretical capacity of Si at 3579 mAh/g. Multiple samples for each electrode formulation were tested. Cell to cell variation is less than 5%. The delithiated electrodes were extracted from cycled cells and carefully rinsed with DMC four times (1 mL in total) to remove residual electrolyte and then dried in a glovebox for ex situ analysis. The infrared spectra with attenuated total reflectance (IR-ATR) analysis was conducted inside an N2-filled glovebox with a Bruker Tensor 27 equipped with an LaDTG detector. All spectra were collected with 512 scans and spectral resolution of 4 cm−1. Ex situ XPS was conducted using a Kα spectrometer (Thermo scientific) with spot size of 400 μm, an energy step size of 0.05 eV, and a pass energy of 50 eV. The electrodes were transferred from the glovebox to the XPS chamber using a sealed transfer apparatus without exposure to the air. The binding energy was corrected on the basis of the C 1s of hydrocarbon at 285 eV. The change in surface morphology of electrodes before and after cycling was examined by ex situ SEM (Sigma VP, ZEISS) at 5 kV.

The use of water as solvent for electrode processing with water-soluble binders such as PAA and CMC reduces production cost and is more environmentally friendly than NMP. However, silicon nanoparticles readily react with water to produce hydrogen and silicon oxide.6,16 If a thick layer of silicon surface oxide is formed on the silicon particle, the electrode conductivity will be decreased along with the capacity and cycling performance.17,18 Thus, the oxidation of silicon should be minimized during electrode processing. Lithium ion cells are frequently stored for 24 h after filling with electrolyte to ensure electrode wetting.19−21 The temperature of the cells is typically maintained slightly higher than room temperature to accelerate the electrolyte penetration process.19,20 While LiPF6 is the most common salt used in commercial batteries, LiPF6 is thermally unstable and reacts with protic impurities such as water.22 Common formulations of Si electrodes with good performance consist of 30−60% Si, 15−30% super C, and 15−30% binders such as PAA and CMC.4,8,14,15 Since PAA and CMC contain high concentrations of carboxylic acid (−COOH) and hydroxyl (−OH) substituents, PAA and CMC likely accelerate the hydrolytic decomposition of the electrolyte especially at elevated temperature. Therefore, understanding the stability of the electrolyte in the presence of silicon electrodes with protic binders is necessary to understand the role of binders in performance enhancement of silicon electrodes. In this manuscript, the interaction of four different binders, PAA, CMC, cross-linked PAA−CMC (c−PAA−CMC), and conventional PVDF, have been investigated with silicon nanoparticles during electrode preparation. In addition, the interaction of binders with electrolytes, the effect of binders on cycling performance, and the effect of binders on SEI formation on silicon nanoparticle anodes has been investigated with electrochemical galvanostatic cycling, attenuated total reflection infrared spectroscopy (ATR-IR), X-ray photoelectron spectroscopy (XPS), and field emission scanning electron microscopy (FE-SEM).

3. RESULTS AND DISCUSSION 3.1. Electrochemical Properties. Figure 1a,b shows a voltage profile and dQ/dV plots for the first cycle of Si anodes

2. EXPERIMENTAL SECTION Poly(acrylic acid) (PAA, Mw = 450 000) and sodium carboxymethyl cellulose (CMC, Mw = 700 000) were purchased from Sigma−Aldrich and used as received. PVDF (Mw = 600 000) was obtained from MTI. Battery grade ethylene carbonate (EC), diethyl carbonate (DEC), dimethyl carbonate (DMC), and lithium hexafluorophospate (LiPF6) were obtained from BASF. Silicon nanoparticles (≤50 nm) were purchased from Alfar Aesar. Conductive carbon (super C) was kindly provided by Timcal (Belgium). Silicon nanoparticles, super C, and binder with a ratio of 50:25:25 were thoroughly mixed in deionized water with PAA, CMC, or their mixture or mixed in NMP with PVDF to prepare a slurry. The slurry was spread on a copper foil and dried in the air at room temperature for 1 h for water or in a convection oven at 60 °C for NMP. The electrodes were punched into 12.7 mm diameter disks and dried in a vacuum oven at 30, 80, and 110 °C overnight and then 150 °C for 4 h. The dry electrodes were not calendared. The thickness of the electrode laminates were ∼15 μm (excluding copper foil), and the total material loading was ∼1.2 mg/cm2 (0.6 mg/cm2 for Si), corresponding to a density of ∼0.8 g/cm−3. 2032 coin cells were assembled in an Ar-filled glovebox and used for evaluation of electrochemical cycling performance. The cells consist of a Si working electrode, a lithium foil counter electrode, electrolyte (100 μL), and a separator (Celgard 2325). The electrolyte was 1.2 M LiPF6 in ethylene carbonate (EC)/diethylene carbonate (DEC) (1:1, w/w). The cells were cycled between 0.005 and 1.5 V at a rate of C/20 (0.179 A/g) for the first cycle (formation cycle) and then C/5 (0.716

Figure 1. Voltage profiles (a) and dQ/dV plots (b) for the first cycle and capacity (c) and efficiency (d) vs cycle of Si anodes with different binders. The dQ/dV was plotted offset for clarify.

with different binders. The curve shape in the dQ/dV plots (Figure 1b) is different for each binder while the silicon and Super C content are constant, suggesting that the binders alter the initial interfacial reactions. The Si−PVDF electrodes contain sharp peaks at 1.3 and 0.7 V resulting from the reduction of electrolyte.23,24 These peaks are modified for electrodes containing PAA and CMC. The Si−PAA electrode exhibits a small peak at 1.6 V, which likely results from the reduction of the −COOH group from PAA binder, and broad peaks at 0.9 and 0.5 V from electrolyte reduction. The Si− CMC electrode contains peaks at 0.55 and 0.23 V, which likely result from reduction of both the electrolyte and hydroxyl 12212

DOI: 10.1021/acsami.6b03357 ACS Appl. Mater. Interfaces 2016, 8, 12211−12220

Research Article

ACS Applied Materials & Interfaces

Figure 2. Low magnification SEM images of Si electrodes with PVDF, PAA, CMC, and PAA−CMC before and after cycling.

no significant changes for Si−PVDF during cycling, probably due to the weak interaction of PVDF with Si nanoparticles. Though macrochanges in the electrode morphology during cycling obviously depend on the chemistry of the binders, the appearance of electrode cracks seems not to be a main contribution to cycling performance of silicon anodes as shown in Figure 1. The differences should relate to the changes in the nanoscale. High magnification SEM images of Si−PVdF and Si−PAA before and after 5 cycles are provided in Figure 3. Both fresh

groups from CMC. The c−Si−PAA−CMC has similar reduction peaks to those observed for the PAA binder. Figure 1c,d exhibits capacity and efficiency versus cycle number of Si anodes with different binders. Both silicon and conductive carbon, i.e., super C, could contribute to the total capacity of the electrodes. However, conductive carbon has very low reversible capacity, about 180 mAh/g25 compared to 3579 mAh/g for silicon. Thus, the specific capacity of silicon electrodes is primarily from silicon. Si−PVDF electrode has an initial capacity of ∼3000 mAg/h based on weight of silicon and a first cycle efficiency of 75%. The specific capacity quickly drops to 1000 mAh/g after 10 cycles with low efficiency. On the contrary, Si electrodes containing PAA, CMC, or c−PAA− CMC have remarkable improvement in both specific capacity and efficiency with first capacity and efficiency of ∼3600 mAh/ g and ∼87%, respectively. The improved first cycle efficiency is consistent with the loss of the electrolyte reduction peaks observed in the dQ/dV plots. After 20 cycles, the capacity of the electrodes still remains ≥3000 mAh/g. Si−PAA anode has similar capacity retention and efficiency to c−Si−PAA−CMC. Interestingly, both electrodes, containing PAA, Si−PAA, and c−Si−PAA−CMC, have slightly better capacity retention than electrodes containing only CMC. 3.2. Electrode Morphology Changes. Low magnification SEM images of the electrodes with different binders before and after cycling are depicted in Figure 2. All fresh electrodes show similar surface morphology, having a uniform smooth surface. After the first cycle, the large cracks appear on Si electrodes containing PAA and CMC. As cycle number increases to 5 and 10, both the number of cracks and crack size increase. After 20 cycles, the number of large cracks somewhat decreases. In addition, electrodes containing PAA, i.e., Si−PAA and Si− PAA−CMC, have more cracks and larger crack size compared to Si−CMC at all cycle numbers. PAA seems to integrate with silicon nanoparticles stronger than CMC. As a result, the Si− PAA has more rigid electrode structure which is easily broken by the large volume changes of silicon. Interestingly, there are

Figure 3. High magnification SEM images of Si−PVdF (a and c) and Si−PAA (b and d) before and after 5 cycles, respectively.

electrodes consist of round-shaped silicon particles with a diameter of 50−100 nm surrounded by super C nanoparticles (Figure 3a and b). After 5 cycles, the Si nanoparticles in the Si− PVdF electrode (Figure 3c) are covered by a thick SEI making the boundary of particles difficult to distinguish. The thick SEI may isolate the Si particles resulting in a rapid capacity fade as 12213

DOI: 10.1021/acsami.6b03357 ACS Appl. Mater. Interfaces 2016, 8, 12211−12220

Research Article

ACS Applied Materials & Interfaces seen in Figure 1. On the contrary, the Si−PAA (Figure 3d) reveals clear individual particles with clean edges. The thinner SEI on Si−PAA ensures the good electric connection between particles during cycling, leading to better cyclability. The SEM images of cycled Si−CMC and Si−PAA−CMC electrodes are similar to the SEM images of cycled Si−PAA electrodes. 3.3. Interaction of Binders with Silicon Anodes during Preparation. The Si 2p spectra for the electrodes with different binders after drying at 150 °C are provided in Figure 4a. Variation of the drying temperature does not result in

Figure 5. ATR FT-IR of Si anode prepared with different binders and drying temperature.

at 80 °C contains a strong peak at 1708 cm−1, characteristic of the −COOH group in PAA. In addition, a new peak is observed at 1610 cm−1, which is attributed to hydrogen bonding carboxylic dimers.31 After drying at a higher temperature, i.e., 150 °C, an additional small, new peak appears at 1803 cm−1, which is characteristic of an anhydride, −OCOOC−, resulting from the condensation reaction of two carboxylic groups and the loss of H2O as shown in eq 1:31

Figure 4. Si 2p spectra of fresh Si anodes prepared with (a) different binders and (b) pure PAA (pH ∼ 2 to 3), PAALi0.7 (pH ∼ 6), and PAALi0.9 (pH ∼ 7.5). All spectra were normalized to the peak of elemental Si at 99.3 eV.

significant changes in the Si 2p XPS spectra. The spectra have been normalized to the peak for elemental silicon in the Si 2p spectrum. All electrodes have common features with a strong peak at 99.3 eV from Si and weaker peak at ∼103.5 eV from a surface silicon oxide, SiOx.26 The reaction of water with silicon can result in the generation of SiO2 or even cause the dissolution of silicon under basic conditions, so we may expect that the electrode prepared in water would have more surface oxide.27,28 However, the peak for SiOx for the electrodes containing PAA or Si−PAA−CMC are similar to that of PVDF. On the contrary, electrodes prepared with CMC show greater intensity of the SiOx peak, consistent with more silicon oxide on the surface. Interestingly, the pH of a diluted solution of CMC is ∼7 while the pH of PAA and PAA−CMC solutions is ∼3. Thus, generation of surface oxide on silicon nanoparticles may be dependent upon the pH of the binder solution. Thus, the oxidation of Si was monitored for electrodes with PAA at different pH values by neutralizing the PAA solution with LiOH (denoted as PAALi0.7 (pH ∼ 6) and PAALi0.9 (pH ∼ 7.5) where 0.7 and 0.9 are mole ratio of LiOH/PAA). As shown in Figure 4b, the Si 2p spectra of the electrodes prepared at different pH values confirm that SiOx increases with increasing pH. Although the cyclability of Si is influenced by adhesion of binders, mechanical strength, dispersion ability of active materials, and porosity of the electrodes,15,29 the formation of thicker silicon oxide layer for Si−CMC likely contributes to the lower performance compared to PAA or PAA−CMC. Figure 5 displays ATR FTIR spectra for Si electrodes dried at 80 and 150 °C. IR spectra for pure binders are also included for reference. The Si electrodes prepared with PVDF show similar spectral features upon drying at either 80 or 150 °C and are dominated by signals originating from PVDF at 1402, 1175, 878, and 840 cm−1.30 The spectrum of Si−PAA electrodes dried

Upon drying the Si−CMC anode at 80 °C, a strong absorbance is observed at 1594 cm−1 from the stretching vibration of −COONa groups of the CMC binder. A new peak is observed at 1728 cm−1, attributed to an ester11 which is likely formed by the reaction of −COONa with silicon oxide or silanol groups at the surface of silicon nanoparticles. The electrodes with both PAA and CMC have changes consistent with a combination of the Si−PAA and the Si−CMC electrodes. The spectra of the electrodes stored at 80 °C are consistent with a combination of −COOH and −COO−. The Si−PAA−CMC electrode dried at 150 °C also shows the presence of anhydride at 1803 cm−1, similar to Si−PAA. The peaks associated with ester formation, as observed for Si-CMC at 1728 cm−1, may be present as a shoulder of the peak at 1718 cm−1. 3.4. Interaction of Electrodes with Electrolytes. In order to investigate the reactivity of the electrodes with electrolyte components, Si electrodes were incorporated into coin cells containing 2 separators (Celgard 2325) and 100 μL of electrolyte, 1.2 M LiPF6 in EC/DEC (1:1, w/w). The cells were stored at 25 °C for 1 or 3 days or at 55 °C for 1 day to simulate cell wetting processes used in industry. The Si 2p and F 1s spectra for Si electrodes after storage in electrolyte at 25 and 55 °C are provided in Figures 6 and 7. The Si 2p spectra have been normalized to the peak for 12214

DOI: 10.1021/acsami.6b03357 ACS Appl. Mater. Interfaces 2016, 8, 12211−12220

Research Article

ACS Applied Materials & Interfaces

species.32 The surface SiOx appears to react with the decomposition products of LiPF6, PF5, or HF which is always present in the electrolyte at tens of ppm level, to form SiOyFz. The Si 2p spectrum of Si−PAA also does not show any change upon storage in the electrolyte at 25 °C for 1 day, but after 3 days, the intensity for the peak for SiOx is significantly decreased and a small peak characteristic of SiOyFz is observed. Upon storage in the electrolyte for 1 day at 55 °C, a greater reduction of the peaks characteristic of SiOx is observed with the related appearance of a small peak associated with SiOyFz. The increased loss of SiOx for the Si−PAA electrode is likely related to the presence of acidic carboxylic acid functional groups which accelerate the thermal decomposition of LiPF6. This process is relatively slow at room temperature but much faster at moderately elevated temperature. Storage of the Si− CMC electrode at 25 or 55 °C does not result in a significant change to the Si 2p spectra, suggesting that the basic −CO2Na group may scavenge HF, forming a weaker carboxylic group −CO2H and insoluble salt NaF: −COONa + HF → − COOH + NaF

(2)

The Si 2p spectra of the Si−PAA−CMC electrode are unchanged upon storage at 25 °C for either 1 or 3 days which is similar to that observed for the Si−CMC electrode. However, after storage in the electrolyte at 55 °C for 1 day, the peak for SiOx decreases and a new peak characteristic of SiOyFz is observed. This suggests that the basic −CO2Na groups can inhibit HF generation upon storage at 25 °C but the ability to inhibit HF generation is limited at 55 °C. In the F 1s spectra (Figure 7), Si−PVDF shows a strong peak at ∼688 eV, attributed to C−F bonds from the PVDF binder and there is no significant change during storage. The fresh Si− PAA electrode contains no F 1s peak, but upon storage in the presence of electrolyte at 25 °C for 1 day, a peak is observed at 687.8 eV characteristic of the decomposition product of LiPF6, LixPFyOz.33 The intensity of the peak at 687.8 eV increases with increased storage time or increased temperature, consistent with further decomposition of LiPF6. The fresh Si−CMC electrode also has no F 1s peak, but upon exposure to the electrolyte for 1 day at 25 °C, a broad peak at 685−688 eV is observed, characteristic of a mixture of LixPFyOz and LiF, respectively. After storage at 25 °C for 3 days or at 55 °C for 1 day, the intensity of the broad peak is increased which is consistent with the generation of more LixPFyOz and LiF. Similarly, upon storage in the presence of electrolyte, the Si− PAA−CMC electrode has changes of the F 1s spectra characteristic of the decomposition of LiPF6 to generate LixPFyOz. The changes support LiPF6 decomposition during storage in the presence of PAA or CMC binders. ATR FT-IR spectra of Si electrodes before and after storage with electrolyte are presented in Figure 8. The IR spectra of Si−PVDF do not show any significant changes upon storage at all conditions investigated. On the other hand, electrodes containing PAA and CMC show strong peaks in the region of 850−700 cm−1 after storage, consistent with the generation of LixPFyOz34,35 related to the decomposition of LiPF6, as observed in the XPS spectra above. The intensity of the peak at 1806 cm−1 has increased in a manner consistent with additional anhydride formation upon storage at 55 °C for Si− PAA and Si−PAA−CMC electrodes. The formation of anhydride without catalyst occurs when electrodes containing PAA are heated at a temperature as high as 150 °C.8,31,36 The formation of anhydride at low temperature, i.e, at 55 °C, is

Figure 6. Si 2p spectra for Si electrodes before and after storage. All spectra were normalized to the peak of elemental Si at 99.3 eV.

Figure 7. F 1s spectra for Si electrodes before and after storage.

elemental silicon in the Si 2p spectrum due to a small variation in peak intensity from sample to sample. The Si 2p spectrum (Figure 6) of the Si−PVDF electrode after storage at 25 °C for 1 day is very similar to the fresh electrode consistent with no reaction of the electrolyte with the electrode. However, upon storage at 25 °C for 3 days or at 55 °C for 1 day, the peak intensity of SiOx at 103.5 eV becomes weaker and a new small peak is observed at ∼106.5 eV, characteristic of SiOyFz 12215

DOI: 10.1021/acsami.6b03357 ACS Appl. Mater. Interfaces 2016, 8, 12211−12220

Research Article

ACS Applied Materials & Interfaces

Figure 8. IR spectra of Si electrode before and after storage.

Figure 9. IR spectra of Si electrode before and after cycling.

the peaks for carboxylate at 1594 and 1426 cm−1 with a related increase in the peak characteristic of esters or carboxylic acid at 1734 cm−1. The change is especially apparent for the electrodes

likely due to the presence of decomposition products of LiPF6 such as PF5 and HF which acts as catalyst. The spectra of Si− CMC stored at 25 °C show a gradual decrease in intensity of 12216

DOI: 10.1021/acsami.6b03357 ACS Appl. Mater. Interfaces 2016, 8, 12211−12220

Research Article

ACS Applied Materials & Interfaces stored at 55 °C. This is consistent with the conversion of carboxylate to carboxylic acid according to eq 2. A similar tendency is also observed at Si−PAA−CMC electrode. 3.5. Effect of Binders on the Formation of the SEI Layer at Si Anodes During Cycling. The IR spectra for fresh and cycled electrodes with different binders are provided in Figure 9. The IR spectra of electrodes lithiated to 0.5 V and then delithiated to 1.5 have been included to provide a better understanding about the role of binder on the formation of the SEI on Si electrodes. The IR spectrum of PAALi independently prepared by reaction of PAA and LiOH with molar ratio of 1:1 is also provided. Upon lithiating the Si−PVDF electrode to 0.5 V, new peaks are observed at 1659 and 1328 cm−1, characteristic of lithium alkyl carbonate ROCOOLi37−40 and at 1488 cm−1 characteristic of Li2CO3. These compounds are similar to those reported on graphite electrodes.41 The intensity of the absorptions of the electrolyte decomposition products is comparable to the intensity of the absorptions associated with the PVDF binder, suggesting the formation of a thick SEI. Alternatively, the electrodes with PAA, CMC, and c−PAA−CMC do not contain absorptions characteristic of ROCO2Li or Li2CO3. The Si− PAA electrodes show a significant difference in the IR spectra: The intensity of peaks characteristic of the carboxylic acid −COOH at 1706 cm−1 are significantly decreased while new peaks consistent with the generation of lithium carboxylates (−COOLi) at 1574 cm−1 are observed. The new absorptions are very similar to that of PAALi. This new layer functions as an SEI and suppresses electrochemical decomposition of electrolyte. The conversion of PAA to PAALi is likely due to electrochemical reduction of carboxylic functionality as follows: −COOH + Li+ + e → −COOLi + 1/2H 2↑

Figure 10. C 1s spectra for Si with different binders during cycling.

(3)

The IR spectra of the Si−CMC electrodes reveal a similar inhibition of electrolyte reduction upon lithiating the cell to 0.5 V while the IR spectra of the Si−PAA−CMC electrodes are consistent with a combination of the Si−PAA and Si−CMC electrodes. Upon additional cycling (1−10 cycles), the intensity of the IR absorptions for ROCOOLi, Li2CO3, and RCOOLi on the Si−PVDF electrode significantly increase, consistent with a thickening of the SEI from the continuous decomposition of the electrolyte. After 10 cycles, the IR spectrum is dominated by the presence of Li2CO3. In the spectrum of Si−PAA electrode, the absorption of the carboxylic acid at 1709 cm−1 is absent after the first cycle, suggesting that the PAA binder is completely converted to PAALi. As the cycle number increases above 5, absorptions for ROCOOLi at 1655 cm−1 and Li2CO3 at 1489 cm−1 are observed but have comparable intensity to that of the PAALi from the reduction of PAA. A similar trend is also observed with Si−CMC and Si−PAA−CMC electrodes. PAA, CMC, and their reduction products effectively suppress the reduction of electrolyte. This is consistent with the observation from SEM above. The C 1s, F 1s, and P 2p XPS spectra of the Si−PVDF, Si− PAA, Si−CMC, and c−Si−PAA−CMC electrodes are presented in Figures 10, 11, and 12, respectively. Upon lithiating the Si−PVDF electrode to 0.5 V, the C 1s spectrum is changed. Two new peaks are observed at 289.9 and 286.5 eV, characteristic of ROCOOLi and/or Li2CO3.38,39 The signal of PVDF at 291 eV as seen in the fresh electrode is covered, indicating the formation of a thick SEI which is consistent with the IR data. With further cycling (1−10 cycles), the spectra of

Figure 11. F 1s spectra for Si with different binders during cycling.

the Si−PVDF electrodes are dominated by a strong peak at ∼289.9 eV, suggesting that the electrode is predominantly covered by lithium alkyl carbonates and Li2CO3. The intensity of the peaks at 291−288 eV also increase for Si electrodes 12217

DOI: 10.1021/acsami.6b03357 ACS Appl. Mater. Interfaces 2016, 8, 12211−12220

Research Article

ACS Applied Materials & Interfaces

Si−CMC, and Si−PAA−CMC, a P 2p peak is observed at 134.5 eV consistent with the generation of LixPFyOz during the first cycle, but no P 2p signal is observed on further cycling suggesting that these compounds are only present in the inner SEI. Interestingly, no P 2p signal is observed at any point during the cycling of the Si−PVDF electrodes. The decomposition of LiPF6 appears to be dominated by reactions with the PAA and CMC binders at the beginning of cycling, likely resulting from reactions of LiPF6 with carboxylic and hydroxyl groups from PAA and CMC binders. The rate of these reactions are negligible after the first cycle because of the absence of carboxylic and hydroxyl groups which are completely electrochemically reduced.

4. CONCLUSIONS The relationship between cyclability, SEI formation, and structure of polymer binders for silicon nanoparticle electrodes has been investigated. The capacity retention of Si−PAA and Si−PAA−CMC is slightly better than the Si−CMC but is much better than Si−PVDF. Electrodes composed of Si−PAA, Si− CMC, and Si−PAA−CMC retain a specific capacity ≥3000 mAh/g after 20 cycles while electrodes composed of Si−PVdF have rapid capacity fade to 1000 mAh/g after 10 cycles. Surface analysis of the electrodes via IR and XPS reveals that PAA and CMC interact strongly with silicon nanoparticles during elctrode preparation and uniformly cover the surface of the silicon particles. In addition, the oxidation of Si during electrode preparation in water is dependent upon the pH of binder solution. The Si−CMC electrode has a thicker silicon oxide layer than Si−PAA and Si−PAA−CMC due to the higher pH of CMC solution. The thick SiOx layer likely contributes to lower performance of the Si−CMC electrode. The reactivity of the electrolyte upon storage with the Si−binder electrodes has also been investigated. PAA accelerates the decomposition of LiPF6, resulting in dissolution of surface silicon oxide and the formation of LiF and LixPFyOz. This process is relatively slow at room temperature but much faster at moderately elevated temperature (55 °C). Interestingly, the surface oxide on the Si−CMC electrode is not significantly altered upon storage, suggesting that the Lewis basic −CO2Na scavenges HF in electrolyte. PAA is electrochemically reduced during the first cycle, converting the carboxylic acids to lithium carboxylates and thus generating a protective layer on the surface of silicon. The binder covered silicon surface layer effectively suppresses the reduction of the carbonate solvents (EC and DEC) resulting in a thinner more stable SEI. After the first cycle, the Si−PAA electrode contains higher concentrations of LiF and LixPFyOz than are observed for the Si−PVDF electrode, suggesting higher reactivity of PAA with LiPF6. The Si−CMC and Si−PAA−CMC electrodes have similar behavior to the Si− PAA electrodes. On the contrary, the electrolyte is continiously reduced on the surface of the Si−PVDF electrode forming a thick SEI resulting in active particle isolation. The Si−PVdF electrodes have rapid capacity fade. The improved performance of Si−PAA, Si−CMC−, and Si−PAA−CMC electrodes is due to the formation of a binder-based SEI generated during electrode preparation and initial cycling which protects the surface for continuous reaction with the carbonate solvents.

Figure 12. P 2p spectra for Si with different binders during cycling.

containing PAA and CMC, but the increase is less pronounced upon lithiating the electrode to 0.5 V compared to the Si− PVDF electrode. The reduction products of the binder suppress electrolyte reduction, consistent with the IR data. However, upon additional cycling (5−20 cycles), the XPS spectra of Si−PAA electrodes contain strong peaks at both 289.9 and 286.5 eV, suggesting the outer SEI is dominated by the presence of lithium alkyl carbonate and Li2CO3. The C 1s XPS spectra of the Si−CMC electrode and Si−PAA−CMC are similar indicating that the presence of CMC or PAA binders inhibit electrolyte reduction on the first cycle due to the presence of binder reduction products passivating the silicon surface. Upon additional cycling, the XPS spectra of the electrode surfaces are dominated by peaks associated with lithium alkyl carbonates and Li2CO3 in the outer SEI. These results are in agreement with the IR observation. The F 1s spectrum of fresh Si−PVDF electrodes contains a strong peak at 688 eV, characteristic of C−F from the PVDF binder (Figure 11). Upon lithiating to 0.5 V, a small new peak appears at 685 eV due to LiF which is formed from the thermal and/or electrochemical decomposition of LiPF6.22,33,42 Upon cycling the Si−PVDF electrode, the intensity of the peak for PVDF decreases in intensity due to the thickening of the SEI. The peak for LiF increases initially but then gradually decreases. Electrodes containing PAA and CMC contain peaks at 685 eV from LiF and a peak at ∼687.3 eV from LixPFyOz. During the first cycle, the peaks for both LiF and LixPFyOz become significant. The intensity of both peaks decrease upon additional cycling (cycles 5−10) suggesting that the LiPF6 reacts with the Si−PAA and Si−CMC electrode surfaces initially, but these reactions are inhibited as the SEI grows. Interestingly, as the cycling continues further (20 cycles), a significant increase in the intensity of the peak associated with LiF is observed. The evolution of the P 2p spectra (Figure 12) is in general agreement with the F 1s spectra. For the Si−PAA,



AUTHOR INFORMATION

Corresponding Author

*Phone: 1-401-874-5071. Fax: 1-401-874-5072. E-mail: [email protected]. 12218

DOI: 10.1021/acsami.6b03357 ACS Appl. Mater. Interfaces 2016, 8, 12211−12220

Research Article

ACS Applied Materials & Interfaces Notes

(17) Xun, S.; Song, X.; Grass, M. E.; Roseguo, D. K.; Liu, Z.; Battaglia, V. S.; Liu, G. Improved Initial Performance of Si Nanoparticles by Surface Oxide Reduction for Lithium-Ion Battery Application. Electrochem. Solid-State Lett. 2011, 14 (5), A61−A63. (18) Xun, S.; Song, X.; Wang, L.; Grass, M. E.; Liu, Z.; Battaglia, V. S.; Liu, G. The Effects of Native Oxide Surface Layer on the Electrochemical Performance of Si Nanoparticle-Based Electrodes. J. Electrochem. Soc. 2011, 158 (12), A1260−A1266. (19) Wang, D. Y.; Xia, J.; Ma, L.; Nelson, K. J.; Harlow, J. E.; Xiong, D.; Downie, L. E.; Petibon, R.; Burns, J. C.; Xiao, a.; Lamanna, W. M.; Dahn, J. R. A Systematic Study of Electrolyte Additives in Li[Ni1/ 3Mn1/3Co1/3]O2 (NMC)/Graphite Pouch Cells. J. Electrochem. Soc. 2014, 161 (12), A1818−A1827. (20) Burns, J. C.; Petibon, R.; Nelson, K. J.; Sinha, N. N.; Kassam, A.; Way, B. M.; Dahn, J. R. Studies of the Effect of Varying Vinylene Carbonate (VC) Content in Lithium Ion Cells on Cycling Performance and Cell Impedance. J. Electrochem. Soc. 2013, 160 (10), A1668−A1674. (21) Trask, S. E.; Li, Y.; Kubal, J. J.; Bettge, M.; Polzin, B. J.; Zhu, Y.; Jansen, A. N.; Abraham, D. P. From Coin Cells to 400 mAh Pouch Cells: Enhancing Performance of High-Capacity Lithium-Ion Cells via Modifications in Electrode Constitution and Fabrication. J. Power Sources 2014, 259, 233−244. (22) Campion, C. L.; Li, W.; Lucht, B. L. Thermal Decomposition of LiPF6-Based Electrolytes for Lithium-Ion Batteries. J. Electrochem. Soc. 2005, 152 (12), A2327−A2334. (23) Zhang, X.; Kostecki, R.; Richardson, T. J.; Pugh, J. K.; Ross, P. N. Electrochemical and Infrared Studies of the Reduction of Organic Carbonates. J. Electrochem. Soc. 2001, 148 (12), A1341−A1345. (24) Zhuang, G. V.; Yang, H.; Blizanac, B.; Ross, P. N. A Study of Electrochemical Reduction of Ethylene and Propylene Carbonate Electrolytes on Graphite Using ATR-FTIR Spectroscopy. Electrochem. Solid-State Lett. 2005, 8 (9), A441−A445. (25) Fransson, L.; Eriksson, T.; Edström, K.; Gustafsson, T.; Thomas, J. Influence of Carbon Black and Binder on Li-Ion Batteries. J. Power Sources 2001, 101 (1), 1−9. (26) Moulder, J. F.; Chastain, J. Handbook of X-Ray Photoelectron Spectroscopy: A Reference Book of Standard Spectra for Identification and Interpretation of XPS Data; Physical Electronics Division, Perkin-Elmer Corp.: Eden Prairie, Minn., 1992. (27) Finne, R. M.; Klein, D. L. A Water-Amine-Complexing Agent System for Etching Silicon. J. Electrochem. Soc. 1967, 114 (9), 965− 970. (28) Seidel, H.; Csepregi, L.; Heuberger, A.; Baumgärtel, H. Anisotropic Etching of Crystalline Silicon in Alkaline Solutions I. Orientation Dependence and Behavior of Passivation Layers. J. Electrochem. Soc. 1990, 137 (11), 3612−3626. (29) Han, Z.-J.; Yabuuchi, N.; Shimomura, K.; Murase, M.; Yui, H.; Komaba, S. High-Capacity Si-Graphite Composite Electrodes with a Self-Formed Porous Structure by a Partially Neutralized Polyacrylate for Li-Ion Batteries. Energy Environ. Sci. 2012, 5 (10), 9014−9020. (30) Bormashenko, Y.; Pogreb, R.; Stanevsky, O.; Bormashenko, E. Vibrational Spectrum of PVDF and Its Interpretation. Polym. Test. 2004, 23 (7), 791−796. (31) Komaba, S.; Shimomura, K.; Yabuuchi, N.; Ozeki, T.; Yui, H.; Konno, K. Study on Polymer Binders for High-Capacity SiO Negative Electrode of Li-Ion Batteries. J. Phys. Chem. C 2011, 115 (27), 13487− 13495. (32) Philippe, B.; Dedryvère, R.; Allouche, J.; Lindgren, F.; Gorgoi, M.; Rensmo, H.; Gonbeau, D.; Edström, K. Nanosilicon Electrodes for Lithium-Ion Batteries: Interfacial Mechanisms Studied by Hard and Soft X-Ray Photoelectron Spectroscopy. Chem. Mater. 2012, 24 (6), 1107−1115. (33) Nadimpalli, S. P. V.; Sethuraman, V. a.; Dalavi, S.; Lucht, B.; Chon, M. J.; Shenoy, V. B.; Guduru, P. R. Quantifying Capacity Loss due to Solid-Electrolyte-Interphase Layer Formation on Silicon Negative Electrodes in Lithium-Ion Batteries. J. Power Sources 2012, 215, 145−151.

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors gratefully acknowledge funding from Department of Energy Office of Basic Energy Sciences EPSCoR Implementation award (DE−SC0007074).



REFERENCES

(1) Obrovac, M. N.; Chevrier, V. L. Alloy Negative Electrodes for LiIon Batteries. Chem. Rev. 2014, 114 (23), 11444−11502. (2) Obrovac, M. N.; Christensen, L. Structural Changes in Silicon Anodes during Lithium Insertion/Extraction. Electrochem. Solid-State Lett. 2004, 7 (5), A93−A96. (3) Ryu, J. H.; Kim, J. W.; Sung, Y.-E.; Oh, S. M. Failure Modes of Silicon Powder Negative Electrode in Lithium Secondary Batteries. Electrochem. Solid-State Lett. 2004, 7 (10), A306−A309. (4) Liu, W.-R.; Yang, M.; Wu, H.-C.; Chiao, S. M.; Wu, N.-L. Enhanced Cycle Life of Si Anode for Li-Ion Batteries by Using Modified Elastomeric Binder. Electrochem. Solid-State Lett. 2005, 8 (2), A100−A100. (5) Li, J.; Lewis, R. B.; Dahn, J. R. Sodium Carboxymethyl Cellulose: A Potential Binder for Si Negative Electrodes for Li-Ion Batteries. Electrochem. Solid-State Lett. 2007, 10 (2), A17−A20. (6) Magasinski, A.; Zdyrko, B.; Kovalenko, I.; Hertzberg, B.; Burtovyy, R.; Huebner, C. F.; Fuller, T. F.; Luzinov, I.; Yushin, G. Toward Efficient Binders for Li-Ion Battery Si-Based Anodes: Polyacrylic Acid. ACS Appl. Mater. Interfaces 2010, 2 (11), 3004−3010. (7) Kovalenko, I.; Zdyrko, B.; Magasinski, A.; Hertzberg, B.; Milicev, Z.; Burtovyy, R.; Luzinov, I.; Yushin, G. A Major Constituent of Brown Algae for Use in High-Capacity Li-Ion Batteries. Science 2011, 334 (6052), 75−79. (8) Koo, B.; Kim, H.; Cho, Y.; Lee, K. T.; Choi, N.-S.; Cho, J. A Highly Cross-Linked Polymeric Binder for High-Performance Silicon Negative Electrodes in Lithium Ion Batteries. Angew. Chem., Int. Ed. 2012, 51 (35), 8762−8767. (9) Buqa, H.; Holzapfel, M.; Krumeich, F.; Veit, C.; Novák, P. Study of Styrene Butadiene Rubber and Sodium Methyl Cellulose as Binder for Negative Electrodes in Lithium-Ion Batteries. J. Power Sources 2006, 161 (1), 617−622. (10) Munao, D.; van Erven, J. W. M.; Valvo, M.; Garcia-Tamayo, E.; Kelder, E. M. Role of the Binder on the Failure Mechanism of Si Nano-Composite Electrodes for Li-Ion Batteries. J. Power Sources 2011, 196 (16), 6695−6702. (11) Vogl, U. S.; Das, P. K.; Weber, A. Z.; Winter, M.; Kostecki, R.; Lux, S. F. Mechanism of Interactions between CMC Binder and Si Single Crystal Facets. Langmuir 2014, 30 (34), 10299−10307. (12) Lestriez, B.; Bahri, S.; Sandu, I.; Roué, L.; Guyomard, D. On the Binding Mechanism of CMC in Si Negative Electrodes for Li-Ion Batteries. Electrochem. Commun. 2007, 9 (12), 2801−2806. (13) Hochgatterer, N. S.; Schweiger, M. R.; Koller, S.; Raimann, P. R.; Wöhrle, T.; Wurm, C.; Winter, M. Silicon/Graphite Composite Electrodes for High-Capacity Anodes: Influence of Binder Chemistry on Cycling Stability. Electrochem. Solid-State Lett. 2008, 11 (5), A76− A76. (14) Bridel, J.-S. S.; Azaïs, T.; Morcrette, M.; Tarascon, J.-M. M.; Larcher, D. Key Parameters Governing the Reversibility of Si/Carbon/ CMC Electrodes for Li-Ion Batteries †. Chem. Mater. 2010, 22 (3), 1229−1241. (15) Jeong, Y. K.; Kwon, T.; Lee, I.; Kim, T.-S.; Coskun, A.; Choi, J. W. Millipede-Inspired Structural Design Principle for High Performance Polysaccharide Binders in Silicon Anodes. Energy Environ. Sci. 2015, 8 (4), 1224−1230. (16) Touidjine, A.; Morcrette, M.; Courty, M.; Davoisne, C.; Lejeune, M.; Mariage, N.; Porcher, W.; Larcher, D. Partially Oxidized Silicon Particles for Stable Aqueous Slurries and Practical Large-Scale Making of Si-Based Electrodes. J. Electrochem. Soc. 2015, 162 (8), A1466−A1475. 12219

DOI: 10.1021/acsami.6b03357 ACS Appl. Mater. Interfaces 2016, 8, 12211−12220

Research Article

ACS Applied Materials & Interfaces (34) Nguyen, C. C.; Song, S.-W. W. Interfacial Structural Stabilization on Amorphous Silicon Anode for Improved Cycling Performance in Lithium-Ion Batteries. Electrochim. Acta 2010, 55 (8), 3026−3033. (35) Socrates, G. Infrared and Raman Characteristic Group Frequencies: Tables and Charts, 3rd ed.; Wiley: Chichester; New York, 2001. (36) Dong, J.; Ozaki, Y.; Nakashima, K. Infrared, Raman, and NearInfrared Spectroscopic Evidence for the Coexistence of Various Hydrogen-Bond Forms in Poly(acrylic Acid). Macromolecules 1997, 30 (4), 1111−1117. (37) Xu, K.; Zhuang, G. V.; Allen, J. L.; Lee, U.; Zhang, S. S.; Ross, P. N.; Jow, T. R. Syntheses and Characterization of Lithium Alkyl Monoand Dicarbonates as Components of Surface Films in Li-Ion Batteries. J. Phys. Chem. B 2006, 110 (15), 7708−7719. (38) Nie, M.; Abraham, D. P.; Chen, Y.; Bose, A.; Lucht, B. L. Silicon Solid Electrolyte Interphase (SEI) of Lithium Ion Battery Characterized by Microscopy and Spectroscopy. J. Phys. Chem. C 2013, 117 (26), 13403−13412. (39) 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 (1), 965−976. (40) Nguyen, C. C.; Lucht, B. L. Comparative Study of Fluoroethylene Carbonate and Vinylene Carbonate for Silicon Anodes in Lithium Ion Batteries. J. Electrochem. Soc. 2014, 161 (12), A1933− A1938. (41) Aurbach, D.; Markovsky, B.; et al. A Comparative Study of Synthetic Graphite and Li Electrodes in Electrolyte Solutions Based on Ethylene Carbonate-Dimethyl Carbonate Mixtures. J. Electrochem. Soc. 1996, 143 (12), 3809−3820. (42) Aurbach, D.; Weissman, I.; Schechter, A.; Cohen, H. X-Ray Photoelectron Spectroscopy Studies of Lithium Surfaces Prepared in Several Important Electrolyte Solutions. A Comparison with Previous Studies by Fourier Transform Infrared Spectroscopy. Langmuir 1996, 12 (16), 3991−4007.

12220

DOI: 10.1021/acsami.6b03357 ACS Appl. Mater. Interfaces 2016, 8, 12211−12220