Influence of a binder on the electrochemical behaviour of Si/RGO composite as negative electrode material for Li-ion batteries 259 Influence of a binder on the electrochemical behaviour of Si/RGO composite as negative electrode material for Li-ion batteries A. V. Korchuna*, E. Yu. Evshchika, S. A. Baskakova , O. V. Bushkovaa,b, Y. A. Dobrovolskya a Institute of Problems of Chemical Physics of the Russian Academy of Sciences, 1 Academician Semenov av., Chernogolovka, Moscow region, 142432, Russia b Institute of Solid State Chemistry of the Ural Branch of the Russian Academy of Sciences, 91 Pervomaiskaya st., Ekaterinburg, Russia *email: andrei_korchun@mail.ru Abstract. A composite consisting of silicon nanoparticles and reduced graphene oxide nanosheets (Si/RGO) was studied as a promising material for the negative elec- trode of lithium-ion batteries. Commonly used polyvinylidene fluoride (PVdF) and car- boxymethyl cellulose (CMC) served as a binder. To reveal the influence of the binder on the electrochemical behaviour of the Si/RGO composite, binder-free electrodes were also prepared and examined. Anode half-cells with composites comprising CMC as a binder demonstrated the best properties: capacity over 1200 mAh·g–1, excellent cycling per- formance and good rate capability up to 1.0C. Keywords: Li-ion battery; negative electrode; nanocomposite; reduced graphene oxide; silicon nanoparticles Received: 02.12.2020. Accepted: 21.12.2020. Published:30.12.2020. © Korchun A. V., Evshchik E. Yu., Baskakov S. A., Bushkova O. V., Dobrovolsky Y. A., 2020 D O I: 1 0. 15 82 6/ ch im te ch .2 02 0. 7. 4. 21 Korchun A. V., Evshchik E. Yu., Baskakov S. A., Bushkova O. V., Dobrovolsky Y. A. Chimica Techno Acta. 2020. Vol. 7, no. 4. P. 259–268. ISSN 2409–5613 Introduction Li-ion batteries (LIBs) are lead- ing electrochemical energy storage systems among secondary batteries due to their high energy density. Graphite, the most common material of the negative electrode, is widely used in LIBs production for several reasons: good cyclability, low cost, non-toxicity, low operating voltage [1]. Commercial gra- phitised materials demonstrate specific ca- pacity near 360 mAh∙g–1 [2], which is very close to the limiting value of the theoretical capacity of graphite, 372 mAh∙g–1. Further improvements in the total capacity of LIBs require a material with a much higher ca- pacity. Silicon is a promising material for the negative electrode due to its high the- oretical capacity of 3579 mAh∙g–1 (which corresponds to the formation of Li3.75Si com- pound [3]); it is almost ten times more than that of graphite. Unfortunately, individual macroscale silicon is not suitable for prac- tical usage in LIBs as a negative electrode because of colossal volume changes during lithiation [4], causing cracks and leading to loss of material integrity [5]. Silicon nano- particles can withstand such volume changes without destruction, but they still have low electronic and ionic conductivity. Moreo- ver, the high surface area of Si nanoparti- cles leads to excessive solid-electrolyte in- 260 terphase (SEI) formation. All these peculiar properties hinder their application in LIBs as the negative electrode material [6]. A  promising way to  incorporate sili- con into LIBs production is combining it with various carbonaceous materials (na- notubes, graphite, amorphous coatings, graphene and others) to obtain a compos- ite [7]. Carbon enhances electronic and ionic transport throughout a  composite and helps maintain material stability [8]. Choice of a binder also plays a significant role in the stability improvement of the sil- icon-containing electrode, providing good adhesion to the current collector and reli- able contact with carbonaceous particles for better electronic conductivity [9, 10]. Reduced graphene oxide seems to be a suitable carbonaceous material to stabi- lize silicon nanoparticles during cycling and provide fast electron transfer in such a composite. Few recent works describe de- sign and modification of Si/RGO compos- ites to achieve better performance during cycling; some researchers tried to modify the silicon particle surface for better con- tact between Si and RGO [11–14]. For ex- ample, covering with poly(diallyl dimeth- ylammonium chloride) which can change the surface charge of silicon to positive was studied [11]. Si/GO composite was assem- bled by electrostatic attraction, and then GO was reduced into RGO by heat treat- ment [11]. Another approach to stabilize the composite structure is to cover Si/RGO particles with pyrolyzed carbon [12, 13]. Other researches also use simple routes to obtain Si/RGO composites with good cycling stability [14]. In  this work, a  new approach to  ob- tain Si/RGO composite was developed, and the influence of a binder on cycling behaviour of the Si/RGO composite was investigated. Experimental Starting materials Graphene oxide water suspension with a concentration of 14.6 mg∙cm–3 was pro- vided by GRAPHENOX LLC (Russia). Sili- con nanoparticles were prepared by the de- struction of monosilane in argon plasma, as described in [15]; average particle size was 50 nm (SEM), and the specific area was 52 m2∙g–1 (by BET). Composite preparation Si/RGO composite with 50:50 ratio (by weight) was prepared by self-assem- bly in a water suspension. 100 cm3 of GO suspension was ultrasonically dispersed for 5  minutes. Then, 774  mg of  silicon nanoparticles were added to the suspen- sion and again ultrasonically dispersed for 5 minutes. After that, the suspension was frozen in  cylindrical moulds with an inner diameter of 20 mm on a copper plate cooled by  liquid nitrogen. Finally, the frozen suspension was freeze-dried for 72 hours at –55 °C (Martin Christ Alpha 1–4 LDplus, Germany). The resulting aero- gel was then reduced in hydrazine vapour atmosphere at room temperature and dried in an oven at 65 °C. Samples thus obtained were named as SiRGO. Electrode preparation Active mass of electrode consisted of Si/ RGO composite and a  binder (PVdF or CMC) at 90:10 ratio (by weight); electrode without binder was also prepared, for com- parison. N-Methyl-2-pyrrolidone (NMP) was used as a solvent for samples with PVdF binder, and water was used as a solvent for samples with CMC binder. PVdF and CMC binders were purchased from Sigma-Al- 261 drich. Table 1 presents the type of polymer binder and the composition of electrode active mass. Samples without binder were designated as SiRGO1, with PVdF bind- er — as SiRGO2, and with CMC binder — as SiRGO3. The area of prepared electrodes was 2.25 cm2. The active material loading of the electrode was about 0.5 mg∙cm–2. Electrode slurry was prepared by adding a suitable solvent to a composite powder with further homogenization by IKA T10 Ultra Turrax equipment (10 min) and then by an ultrasonic bath (10 min). Electrode slurry without binder was prepared by add- ing NMP to a composite powder with subse- quent homogenization as described above. Electrode slurry was applied onto a copper current collector using Dr Blade technology with subsequent calendering and vacuum drying at 120 °C for 12 h. Methods The morphology of the composite sam- ples was studied by scanning electron mi- croscopy (SEM) using Zeiss LEO SUPRA 25 (Germany) equipment with accelerating voltage of 12 kV. E l e c t ro che m i c a l char a c te r ist i c s of the Si/RGO composite electrodes with various binders were measured in  two- electrode pouch cells modelling anode half-element of LIB. The cells were assem- bled in an argon-filled glove box. Lithium metal foil on a  copper current collector served as a counter electrode, and a base- line 1 m solution of  LiPF6 in  ethylene carbonate (EC)/ethyl methyl carbonate (EMC) mixture (1:1 by  volume) served as an electrolyte; Celgard 2300 film served as a separator. The residual water content in the electrolyte solution did not exceed 30 ppm. All components were supplied by Sig- ma-Aldrich. Cyclic voltammetry (CV) and galvanostatic cycling (GST) were per- formed using a multi-channel potentiostat P20X8 (“Elins” LLC, Russia). In CV meas- urements, the potential range was 10–2000 mV with a sweep rate of 0.1 mV/s. GST was performed in  two regimes: 1) with con- stant current throughout the whole cycling (0.1C); 2) with constant charge rate of 0.1C and different discharge rates of 0.1C, 0.2C, 0.3C, 0.5C and 1.0C. Intermediate dis- charge at 0.1C for 10 cycles between every 10 cycles with different discharge rates was used. (0.1C rate corresponded to 0.1 A g–1; this value was estimated from the supposed capacity of the Si/RGO active mass equal to 1000 mAh∙g–1). Results and discussion Scanning electron microscopy Fig.  1 illustrates the  morphology of the Si/RGO composite studied by scan- ning electron microscopy. RGO provides a flexible interconnected matrix in which agglomerated silicon nanoparticles are randomly distributed. SEM data demon- strate that some of  the  agglomerated Si nanoparticles occupy the  surface posi- tions, but most of them allocated between RGO sheets. As mentioned in Introduc- tion, RGO sheets can enhance the stability of Si-based composite material during Li+ insertion-extraction cycles since they pro- vide fast electron transfer and mechanical support for silicon nanoparticles. Voids Table 1 Compositions of electrode active mass Sample Binder Composite:binder ratio SiRGO1 None — SiRGO2 PVdF 90:10 SiRGO3 CMC 90:10 262 in the composite structure allow accom- modating volume changes of  silicon na- noparticles during cycling, thus stabilising the electrode. Cyclic voltammetry CV curves allow a better understand- ing of  the  electrochemical behaviour of the Si/RGO electrodes during cycling. Fig. 2 shows the initial three cycles for all samples under investigation. One can see that anode half-cells comprising Si/RGO electrodes with and without polymer bind- er demonstrated different electrochemi- cal behaviour. For the SiRGO1 electrode without binder cathodic and anodic curves for the 1st, 2nd and 3rd cycle are practically the  same. In  contrast, for the  SiRGO2 and SiRGO3 electrodes with a  polymer binder, well-distinguishable peaks appear only in the second cycle and their inten- sity increases by the third cycle. The posi- tions of the peaks are in good agreement with the literature data [16–19], according to which they can be attributed to the pro- cesses of  insertion/extraction of  lithium into silicon or RGO nanoparticles and to the SEI formation. Gradual appearance of the peaks on cycling (Fig. 2(b, c)) can be attributed to the influence of polymer binder on the surface chemistry of silicon particles distributed in the carbon matrix. CMC can bind with silicon surface by for- mation of ester or hydrogen bonds [9, 10], -500 0 500 1000 1500 2000 2500 3000 3500 -600 -400 -200 0 200 Ь anodic path I, µA E, mV 1st cycle 2nd cycle 3rd cyclecathodic pathЮ A -500 0 500 1000 1500 2000 2500 3000 3500 -600 -400 -200 0 200 cathodic pathЮ I, µ A E, mV 1st cycle 2nd cycle 3rd cycle Ь anodic path B -500 0 500 1000 1500 2000 2500 3000 3500 -1200 -1000 -800 -600 -400 -200 0 200 400 600 Ь anodic path I, µ A E, mV 1st cycle 2nd cycle 3rd cyclecathodic pathЮ C Fig. 2. Cyclic voltammograms of (A) SiRGO1; (B) SiRGO2; (C) SiRGO3 Fig. 1. Si/RGO composite SEM images with different magnifications 263 whereas PVdF forms with silicon surface Van-der-Vaals bonds only. Chemical bonds formation caused by CMC influences on the positions and intensity of the typical peaks. Galvanostatic cycling The  electrochemical performance of  anode half-cells comprising Si/RGO nanocomposites was studied by galvano- static charge-discharge cycling in the range 10 to  2000 mV under the  same condi- tions at  room temperature. Fig.  3 shows the charge-discharge curves in the 1st, 2nd and 5th cycles for the  cells with SiRGO1 (Fig. 3A), SiRGO2 (Fig. 3C) and SiRGO3 (Fig.  3E) electrodes. In  the  first cycle, the initial discharge capacities of the elec- trodes were 573  mAh∙g–1, 1350  mAh∙g–1 and 249 mAh∙g–1, the values of the Cou- lomb efficiency were 56%, 69% and 21%, respectively. Table 2 summarises the elec- trochemical behaviour of electrodes based on Si/RGO nanocomposites with different binders. The  discharge curves of  the  1st cycle are typical for formation cycles in  LIBs. All the curves contain a step near 800 mV, which corresponds to the reduction of elec- trolyte components on the surface of active materials Si and RGO resulting in a protec- tive SEI layer formation. This step disap- peared in the next cycles (Fig. 3). Forma- tion of SEI is the well-known reason for low Coulombic efficiency in the first cycle [7]. The enhanced surface electrochemical reactivity of Si/RGO nanocomposites must be attributed to the large surface-to-volume ratio of both RGO and Si [15, 20]. How- ever, the sample SiRGO3 with CMC binder differ from others: one can see that the first discharge curve contains one more step near 1500 mV. This new process is respon- sible for the largest decrease in Coulomb efficiency at the first cycle (Table 2). Most likely, a polymer binder also participates in the reduction processes during the first cathodic polarisation of the Si/RGO elec- trode. This conclusion is  in  good agree- ment with the CV data. As can be seen from Fig. 3 (B, D, F), the cyclic behaviour of the cell compris- ing SiRGO3 electrode is  fundamentally different from others. Indeed, the  ca- pacity of  electrodes without a  polymer binder (SiRGO1, Fig. 3B) and with PVDF as a binder (SiRGO2, Fig. 3D) drops upon cycling, while the capacity of an electrode with a  CMC binder (Fig.  3E) increases rapidly during the first 8 cycles and then stabilises near 1200 mAh∙g–1 (Table 2). The theoretical capacity of composite samples Si/RGO can be calculated based on the content of Si nanoparticles and RGO in  the  composite using their theoretical capacity. As was mentioned above, Si has the theoretical capacity of 3579 mAh∙g–1 and graphene, by  different evaluations, has the theoretical capacity ranging from 500 to 1116 mAh∙g–1 [21]. The composite contained 50 mass % Si nanoparticles and 50 mass % RGO. The theoretical capacity (Qth) of the composite is calculated using formula (1): ,th Si Si RGO RGOQ Q Q= ⋅ω + ⋅ω (1) where QSi and QRGO are theoretical capaci- ties of Si and RGO, respectively; ωSi and ωRGO are mass fractions of  Si and RGO in the composite, respectively. The resulting value of the theoretical capacity of the Si/ RGO composite, using the minimum value of 500 mAh∙g–1 as the theoretical capacity of RGO, was estimated as 2040 mAh∙g–1. As can be seen from Table 2, the capacity values of all samples under investigation fall below the theoretical one, which is typ- ical for any Si-based anode materials [2–7]. 264 Fig. 3. Charge-discharge curves (A, C, E) and dependences of discharge capacity and Coulombic efficiency (B, D, F) on cycle number for SiRGO1 (A, B), SiRGO2 (C, D) and SiRGO3 (E, F) 265 As follows from the obtained results, CMC binder can solely enhance the stabil- ity of silicon-carbon nanocomposite. This conclusion is in line with those presented in [22]. In [23], authors declare that CMC alone is too brittle to effectively stabilise Si- based negative electrodes during cycling. However, in this work, CMC demonstrated excellent stabilisation properties. The ob- served effect can be attributed to the for- mation of hydrogen bonds between func- tional groups on the  surface of  silicon particles and carboxyl groups in  CMC and between CMC carboxyl groups and residual functional groups in the surface of  RGO nanosheets. Such bonds can be broken during volume expansion of sili- con nanoparticles during lithiation and re- stored during delithiation thus preventing the material from the isolation of particles and helps maintain material integrity [10]. Rate capability was studied for the best SiRGO3 electrode; different discharge rates of  0.1C, 0.2C, 0.3C, 0.5C and 1.0C were used. Fig.  4 presents the  results of  these measurements. The first 10 cycles at 0.1C were used as  the  formation cycles. One can see that the SiRGO3 electrode shows good rate capability up to 1.0C, which can be attributed to both small silicon particle size (50 nm) and presence of conductive RGO layers which support fast electron transfer through the composite and ena- bles high power operation of  half-cell Fig. 4. Rate capability of the SiRGO3 electrode with CMC binder Table 2 The cycle performance of the Si/RGO electrodes in the range of 0.01–2 V at 0.1C rate Sample Discharge capacity, mAh∙g–1 Coulombic efficiency, % Capacity retention, % 1st cycle 60th cycle 100th cycle 1st cycle 60th cycle 100th cycle 60th cycle 100th cycle SiRGO1 573 292 - 56 99 - 51 - SiRGO2 1350 512 334 69 98 99 38 25 SiRGO3 249 1186 1200 21 99 100 100 100 266 [7]. The SiRGO3 electrode demonstrated the capacity of ~1200 mAh∙g–1 at any dis- charge rate with the  exception of  1.0C. At  1.0C, the  electrode capacity reached ~1200 mAh∙g–1 only at the 4th cycle. Per- haps this is due to some diffusion difficul- ties or rearrangement of  the  conducting paths. However, after that, the  capacity again stabilised at the ~1200 mAh∙g–1 val- ue. Some anomalies were also observed at 0.5C, when the capacity first deviated towards higher values causing Coulom- bic efficiency above 100%, and then stabi- lised at the previous level of 1200 mAh∙g–1. The reasons for this behaviour are unclear and require additional examination. Conclusions Polymer binder plays a  crucial role in stabilising the Si/RGO nanocomposite material during cycling. The electrode with PVdF binder demonstrated better initial Coulombic efficiency of 69%, but the ex- ponential capacity loss was observed down to capacity retention 25% after 100 cycles. The electrode with CMC binder exhibited low initial Coulombic efficiency of  21% due to additional reduction process with binder participation. However, the result- ing SEI provided stable cycling with dis- charge capacity near 1200  mAh∙g–1 and 100% capacity retention after 100 cycles at 0.1C. Excellent rate capability up to 1.0C with no capacity fade during cycling was also observed. Acknowledgements This work was performed with financial support from the Ministry of Science and Higher Education of Russian Federation, project ID RFMEFI60419X0235. References 1. Asenbauer J, Eisenmann T, Kuenzel M, Kazzazi A, Chen Z, Bresser D. The success story of graphite as a lithium-ion anode material — fundamentals, remaining chal- lenges, and recent developments including silicon (oxide) composites. Sustain Energy Fuels. 2020;4(11):5387–416. doi:10.1039/D0SE00175A 2. Chae S, Choi S-H, Kim N, Sung J, Cho J. Integration of graphite and silicon anodes for the commercialization of high-energy lithium-ion batteries. Angew Chem Int Ed Engl. 2020;59(1):110–35. doi:10.1002/anie.201902085 3. Obrovac MN, Christensen L. Structural changes in silicon anodes during lithium insertion/extraction. Electrochem Solid State Letters. 2004;7(5):A93. doi:10.1149/1.1652421 4. Luo F, Liu B, Zheng J, Chu G, Zhong K, Li H, et al. Review — nano-silicon/carbon composite anode materials towards practical application for next generation Li-ion batteries. J Electrochem Soc. 2015;162(14):A2509–28. doi:10.1149/2.0131514jes 5. Wu H, Cui Y. Designing nanostructured Si anodes for high energy lithium ion bat- teries. Nano Today. 2012;7(5):414–29. doi:10.1038/35104644 267 6. Su X, Wu Q, Li J, Xiao X, Lott A, Lu W, et al. Silicon-based nanomaterials for lithium- ion batteries: A review. Adv Energy Mater. 2014;4(1):1300882. doi:10.1002/aenm.201300882 7. Shi Q, Zhou J, Ullah S, Yang X, Tokarska K, Trzebicka B, et al. A review of recent developments in Si/C composite materials for Li-ion batteries. Energy Storage Mater. 2021;34:735–54. doi:10.1016/j.ensm.2020.10.026 8. Wu J, Cao Y, Zhao H, Mao J, Guo Z. The critical role of carbon in marrying sili- con and graphite anodes for high-energy lithium-ion batteries. Carbon Energy. 2019;1(1):57–76. doi:10.1002/cey2.2 9. Lestriez B, Bahri S, Sandu I, Roue L, Guyomard D. On the binding mechanism of  CMC in  Si negative electrodes for Li-ion batteries. Electrochem commun. 2007;9(12):2801–6. doi:10.1016/j.elecom.2007.10.001 10. Istomina AS, Bushkova OV. Polimernye svyazuyushchie dlya elektrodov litievykh akkumulyatorov chast’ 1. Polivinilidenftorid, ego proizvodnye i drugie kommert- sializovannye materialy [Polymer binders for the electrodes of lithium batteries. Part 1. Polyvinylidene fluoride, its derivatives and other commercialized materials]. Elektrokhimicheskaya energetika [Electrochemical Energetics]. 2020;20(3):115–131. Russian. doi:10.18500/1608-4039-2020-20-3-115-131 11. Zhou X, Yin Y-X, Wan L-J, Guo Y-G. Self-assembled nanocomposite of silicon na- noparticles encapsulated in graphene through electrostatic attraction for lithium-ion batteries. Adv Energy Mater. 2012;2(9):1086–90. doi:10.1002/aenm.201200158 12. Ding N, Chen Y, Li R, Chen J, Wang C, Li Z, et al. Pomegranate structured C@ pSi/rGO composite as high performance anode materials of lithium-ion batteries. Electrochim Acta. 2020;(137491):137491. doi:10.1016/j.electacta.2020.137491 13. Agyeman DA, Song K, Lee G-H, Park M, Kang Y-M. Carbon-coated Si nanoparticles anchored between reduced graphene oxides as an extremely reversible anode mate- rial for high energy-density Li-ion battery. Adv Energy Mater. 2016;6(20):1600904. doi:10.1002/aenm.201600904 14. Botas C, Carriazo D, Zhang W, Rojo T, Singh G. Silicon-reduced graphene oxide self-standing composites suitable as binder-free anodes for lithium-ion batteries. ACS Appl Mater Interfaces. 2016;8(42):28800–8. doi:10.1021/acsami.6b07910 15. Novikov D. V., Evschik E. Yu., Berestenko V. I., Yaroslavtseva T. V., Levchenko A. V., Kuznetsov M. V., Bukun N. G., Bushkova O. V., Dobrovolsky Yu. A. Electrochemical performance and surface chemistry of nanoparticle Si@SiO2 Li-ion battery anode in LiPF6-based electrolyte. Electrochimica Acta 2016;208:109–119. doi:10.1016/j.electacta.2016.04.179 268 16. Tokur M, Algul H, Ozcan S, Cetinkaya T, Uysal M, Akbulut H. Closing to scaling-up high reversible Si/rGO nanocomposite anodes for lithium ion batteries. Electrochim Acta. 2016;216:312–9. doi:10.1016/j.electacta.2016.09.048 17. Zhai W, Ai Q, Chen L, Wei S, Li D, Zhang L, et al. Walnut-inspired microsized porous silicon/graphene core — shell composites for high-performance lithium-ion battery anodes. Nano Res. 2017;10(12):4274–83. doi:10.1007/s12274-017-1584-5 18. Ogata K, Salager E, Kerr CJ, Fraser AE, Ducati C, Morris AJ, et al. Revealing lithium- silicide phase transformations in nano-structured silicon-based lithium ion batteries via in situ NMR spectroscopy. Nat Commun. 2014;5(1):3217. doi:10.1038/ncomms4217. 19. Feng J, Zhang Z, Ci L, Zhai W, Ai Q, Xiong S. Chemical dealloying synthesis of po- rous silicon anchored by in situ generated graphene sheets as anode material for lithium-ion batteries. J Power Sources. 2015;287:177–83. doi:10.1016/j.jpowsour.2015.04.051 20. Szczech JR, Jin S. Nanostructured silicon for high capacity lithium battery anodes. Energy Environ Sci. 2011;4(1):56–72. doi:10.1039/C0EE00281J 21. Kheirabadi  N., Shafiekhani  A.  Graphene/Li-ion batter y. J Appl Phys. 2012;112(12):124323. doi:10.1063/1.4771923 22. Xiao J, Xu W, Wang D, Choi D, Wang W, Li X, et al. Stabilization of silicon anode for Li-ion batteries. J Electrochem Soc. 2010;157(10):A1047. doi:10.1149/1.3464767 23. Casimir A, Zhang H, Ogoke O, Amine JC, Lu J, Wu G. Silicon-based anodes for lithium-ion batteries: Effectiveness of materials synthesis and electrode preparation. Nano Energy. 2016;27:359–76. doi:10.1016/j.nanoen.2016.07.023