# Fileset

[Carbon Neutralization_2024_Wu.pdf](https://mdr.nims.go.jp/filesets/a3e68fd7-ba3c-415a-a56e-f0ebae6dbd77/download)

## Creator

Gang Wu, Yuanhang Gao, Zheng Weng, Zhicheng Zheng, Wenqiang Fan, Anqiang Pan, Ning Zhang, Xiaohe Liu, [Renzhi Ma](https://orcid.org/0000-0001-7126-2006), Gen Chen

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Binder‐induced inorganic‐rich solid electrolyte interphase and physicochemical dual cross‐linked network for high‐performance SiO<sub>  <i>x</i></sub> anode](https://mdr.nims.go.jp/datasets/b843e533-45ba-44e5-a7b5-4de5d3d0b938)

## Fulltext

Binder‐induced inorganic‐rich solid electrolyte interphase and physicochemical dual cross‐linked network for high‐performance SiOx anodeReceived: 30 April 2024 | Accepted: 14 July 2024DOI: 10.1002/cnl2.158RE S EARCH ART I C L EBinder‐induced inorganic‐rich solid electrolyteinterphase and physicochemical dual cross‐linkednetwork for high‐performance SiOx anodeGang Wu1 | Yuanhang Gao1 | Zheng Weng1 | Zhicheng Zheng1 |Wenqiang Fan1 | Anqiang Pan1 | Ning Zhang1 | Xiaohe Liu1,2 |Renzhi Ma3 | Gen Chen11Key Laboratory of Electronic Packaging and Advanced Functional Materials of Hunan Province, School of Materials Science and Engineering,Changsha, China2School of Chemical Engineering, Zhongyuan Critical Metals Laboratory, Zhengzhou University, Zhengzhou, China3International Center for Materials Nanoarchitectonics (WPI‐MANA), National Institute for Materials Science (NIMS), Tsukuba, JapanCorrespondenceAnqiang Pan, Ning Zhang, Xiaohe Liu andGen Chen, Key Laboratory of ElectronicPackaging and Advanced FunctionalMaterials of Hunan Province, School ofMaterials Science and Engineering,Changsha, Hunan 410083, China.Email: pananqiang@csu.edu.cn, nzhang@csu.edu.cn, liuxiaohe@zzu.edu.cn andgeenchen@csu.edu.cnFunding informationNational Natural Science Foundation ofChina, Grant/Award Number: 22379166;Natural Science Foundation forDistinguished Young Scholars of HunanProvince, Grant/Award Number:2022JJ10089; Key Research andDevelopment Program of Hunan Province,Grant/Award Number: 2023GK2015;Central South University Innovation‐Driven Research Programme,Grant/Award Number: 2023CXQD034AbstractSilicon oxide (SiOx) is heralded as the forefront anode material for high‐energydensity lithium‐ion batteries, owing to its exceptional specific capacity. Never-theless, the traditional combination of polyacrylic acid binder and acetylene blackconductive carbon continues to struggle with the immense stress induced by therepetitive volume expansion and contraction processes. Here we report a highionic conductivity, sulfonyl fluoro‐containing binder for SiOx anode via free radicalcopolymerization reaction between perfluoro (4‐methyl‐3,6‐dioxaoct‐7‐ene) sulfo-nyl fluoride and acrylic acid. The electrode fabrication process incorporatedamino‐functionalized carbon nanotubes (CNT‐NH2) as the conductive agent. Athree‐dimensional conductive network structure is constructed through physicaland chemical double cross‐linking interactions between the ‐COOH and ‐SO2Ffunctional groups of PAF0.1 binder, the ‐NH2 groups of CNT‐NH2, and the ‐OHgroups on the surface of SiOx, including hydrogen bonds and covalent bonds. Inaddition, the binder induces the formation of a solid electrolyte interphase (SEI)rich in inorganic components such as Li2O, Li2SO3, Li2CO3, and LiF. Benefitingfrom the synergistic effects of the physically and chemically double cross‐linkedthree‐dimensional conductive network constructed by the PAF0.1 binder andCNT‐NH2, coupled with the rich‐inorganic SEI, the SiOx anode deliversexceptional rate performance, cycle stability, and lithium‐ion diffusion dynamics.KEYWORD S3D network structure, binder, physicochemical dual crosslinking, SiOxCarbon Neutralization. 2024;3:857–872. onlinelibrary.wiley.com/r/carbon-neutralization | 857This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, providedthe original work is properly cited.© 2024 The Author(s). Carbon Neutralization published by Wenzhou University and John Wiley & Sons Australia, Ltd.Gang Wu and Yuanhang Gao contributed equally to this work.mailto:pananqiang@csu.edu.cnmailto:nzhang@csu.edu.cnmailto:nzhang@csu.edu.cnmailto:liuxiaohe@zzu.edu.cnmailto:geenchen@csu.edu.cnhttps://onlinelibrary.wiley.com/r/carbon-neutralizationhttp://creativecommons.org/licenses/by/4.0/1 | INTRODUCTIONAmidst the intensifying urgency to resolve “low batteryanxiety” in portable electronics, mitigate “range anxiety”in electric vehicles, and tackle “security anxiety” in large‐scale energy storage systems, there exists a ferventanticipation for the emergence of lithium‐ion batteriescharacterized by unparalleled energy density, exceptionalrate performance, and formidable cycling sustainability.[1]However, as the representative of the forefront anodematerial for high‐energy density lithium‐ion batteries,silicon oxide (SiOx) presents difficulties in meeting theurgent commercial demand due to various reasons.[2]Typically, the large volume expansion (~200%), low ionicelectronic conductivity, and irreversible first dischargeproducts (Li2O and LixSiyOz) are three stumbling blocksfor SiOx to realize superior cycle stability, high rateperformance, and enviable initial Coulombic effi-ciency (ICE).[3]Generally, the binder, conductive agent and SiOx arethe main components of the electrode.[4] The interactionamong different components is of vital importance tomaintain the stability of the electrode structure.[5]Traditional polyacrylic acid (PAA) binders formirreversible covalent bonds with the polar groups ofSiOx, making them unable to withstand the large volumefluctuations during cycling.[6] This leads to particlefragmentation and continuous solid electrolyte inter-phase (SEI) rupture, resulting in poor cycle stability andnotably low ICE. Furthermore, traditional acetyleneblack (CB) conductive agents lack polar groups on theirsurface, relying solely on weak van der Waals forces tomaintain electrical contact between SiOx and the currentcollector.[7] Consequently, when SiOx particles undergopulverization due to volume expansion during cycling,the electrical connectivity with the current collector andamong SiOx particles is compromised. The quest forsuitable binders and conductive agents becomes impera-tive to establish robust electrical connections within theelectrode structure.[8]In this work, a high ionic conductivity, sulfonylfluoro‐containing binder, namely PAF binder, for SiOxanode was synthesized via free radical copolymerizationreaction between perfluoro (4‐methyl‐3,6‐dioxaoct‐7‐ene)sulfonyl fluoride (PFSF) and acrylic acid (AA). Theelectrode fabrication process incorporated amino‐functionalized carbon nanotubes (CNT‐NH2) as theconductive agent. A three‐dimensional conductive net-work structure is constructed through physical andchemical double cross‐linking interactions between the‐COOH and ‐SO2F functional groups of PAF0.1 binder,the ‐NH2 groups of CNT, and the ‐OH groups on thesurface of SiOx, including hydrogen bonds and covalentbonds. This intricate network structure effectivelypreserved electrical connectivity among the binder,conductive agent, SiOx particles, and the copper currentcollector. Furthermore, the gradient hydrogen bonds inthe three‐dimensional network structure can graduallydissociate, effectively buffering the stress caused by thevolume expansion of SiOx. Additionally, compared toelectrolyte components, the binder molecules exhibited alower lowest unoccupied molecular orbital (LUMO)energy level. This characteristic made them more proneto preferential reduction on the anode surface duringdischarge, leading to the formation of a SEI composed ofinorganic components like Li2SO3/Li2CO3 and LiF.Benefiting from the synergistic effects of the physicallyand chemically double cross‐linked three‐dimensionalconductive network constructed by the PAF0.1 binderand CNT‐NH2 conductive agent, coupled with the SEIrich in inorganic components, the SiOx anode demon-strated exceptional rate performance, cycle stability, andlithium‐ion diffusion dynamics.2 | RESULTS AND DISCUSSION2.1 | Theoretical calculationTo investigate the interactions between binders andCNT‐NH2, density functional theory (DFT) was em-ployed to calculate the bond energies of various hydrogenbonds and covalent bonds. Previous research indicatesthat the ‐COOH groups in AA and the ‐NH2 groups inCNT‐NH2 can undergo an amidation reaction duringdrying at 150°C, forming covalent amide bonds.[9]Moreover, hydrogen bonds are observed between thepolar ‐COOH and ‐SO2F groups of the PAFx binder andthe ‐NH2 groups of the CNT‐NH2. In Figure 1c, there isan illustration of an S =O ∙ ∙ ∙H‐OOC hydrogen bondwith a bond energy measuring 2.39 kcal mol−1 betweenPFSF and AA molecules. Additionally, there is adepiction of a C =O ∙ ∙ ∙H‐OOC hydrogen bond with ahigher bond energy of 20.08 kcal mol−1 between AAmolecules. Additionally, an S =O ∙ ∙ ∙H‐NH hydrogenbond with a bond energy of 2.86 kcal mol−1 is observedbetween PFSF and CNT‐NH2. Furthermore, there areCOO‐H ∙ ∙ ∙H‐NH hydrogen bonds (2.99 kcal mol−1) andcovalent amide bonds (C =O‐NH) with a significantlyhigher bond energy (271.71 kcal mol−1), linking AA andCNT‐NH2. The observations indicate that the resilientthree‐dimensional framework established by potentcovalent bonds featuring high bond energy between AAand CNT‐NH2 plays a pivotal role in upholding thestructural integrity of the SiOx anode. Additionally, thegradient hydrogen bonds between AA, PFSF, and858 | 27693325, 2024, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cnl2.158 by National Institute For, Wiley Online Library on [21/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fcnl2.158&mode=CNT‐NH2 can be progressively dissociated, effectivelymitigating the stress resulting from the volume expan-sion of SiOx. This comprehensive understanding of bondenergies and interactions aids in designing electrodeswith enhanced stability and performance.To explore in depth the influence of binders on SEIcomponents, DFT was used to continue to evaluate themolecular orbital energy levels of various electrolytes andbinders.[10] Figure 1b illustrates that binder moleculesexhibit lower LUMO energy levels compared to electrolytecomponents like ethylene carbonate (EC), dimethylcarbonate, ethyl methyl carbonate, floroethylene carbonate(FEC), and LiPF6. A lower LUMO energy level signifies ahigher electron affinity, making it easier for the bindermolecules to undergo reduction on the anode surfaceduring the discharge process, leading to the formation ofSEI components. Specifically, the binder molecule with thelowest LUMO energy level suggests that PFSF moleculesenriched with ‐SO2F groups are more inclined to bepreferentially reduced on the anode surface. Consequently,this results in the creation of an SEI that is rich ininorganic components. This insight into molecular orbitalenergy levels aids in understanding the electrochemicalprocesses occurring at the electrode–electrolyte interfaceand underscores the role of binders in influencing SEIcomposition, which is crucial for optimizing batteryperformance and stability.2.2 | Synthesis and characterization ofPAF binderTo unveil the chemical structures of the PFSF, AA, andPAF0.1, Fourier transform infrared spectroscopy (FTIR)analysis was employed. As depicted in Figure 2a, thespectrum of PFSF reveals a minor peak at 1839 cm−1,indicative of the stretching vibration associated with itsC = C double bond. Notably, two distinct peaks at ~1464and 1234 cm−1 emerge, correlating respectively to thesymmetric and asymmetric stretching vibrations of ‐SO2‐.A cluster of minor peaks around 1150 cm−1 is attributedto the stretching vibration of the carbon‐fluorine (C‐F)bond. Furthermore, a sharp peak at 984 cm−1 signifiesthe stretching vibration of the C‐O‐C bond. Noteworthyis the broad peak at 810 cm−1, coupled with a sharp peakat 606 cm−1, indicative of the presence of the S‐F and C‐SFIGURE 1 (a) AA and PFSF monomers are free radically polymerized to synthesize PAF binder. (b) Frontier molecular orbitals ofdifferent compounds. (c) The hydrogen bond types and bond energies calculated by density functional theory simulations. AA, acrylic acid;APS, ammonium persulfate; PFSF, perfluoro (4‐methyl‐3,6‐dioxaoct‐7‐ene) sulfonyl fluoride.| 859 27693325, 2024, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cnl2.158 by National Institute For, Wiley Online Library on [21/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fcnl2.158&mode=bonds, respectively.[11] The spectrum of AA reveals abroad peak spanning from 2751 to 3331 cm−1, indicativeof the hydroxyl stretching vibration within the carboxylgroup (‐OH in ‐COOH).[12] Additionally, the small peakat 1724 cm−1 corresponds to the ‐COO‐ stretchingvibration in the carboxyl group. The absorption peaksobserved at 1693 and 1613 cm−1 are attributed to thestretching vibration of the C =O bond and C=C bond,respectively.[13] In the PAF0.1 copolymer, the character-istic peaks associated with the C = C bonds disappeararound 1839 and 1613 cm−1, whereas new peaks from thesulfonyl fluoride group (‐SO2F) of PFSF and the carboxylgroup (‐COOH) of AA emerge. Compared with PFSFmonomer, C‐F in PAF0.1 undergoes a blue shift due tothe steric hindrance effect between functional groupsafter copolymerization. This provides evidence ofcopolymerization events taking place. Figure 2b depictsthe utilization of X‐ray photoelectron spectroscopy (XPS)analysis to investigate the groups present in PAF0.1. TheF 1s spectrum clearly exhibits two peaks, located at~688.1 and 685.7 eV, corresponding to SO2F and C‐F,respectively, indicating the presence of PFSF in thecopolymer PAF0.1.Given the above results, FTIR was carried out todetect the prevalence of hydrogen bonding in PAFx(x= 0.05 or 0.1). As illustrated in Figure 2c, thedistinctive peak at approximately 1689 cm−1 in the PAAspectrum corresponds to the stretching vibration ofC =O. Conversely, in the spectrum of PAF0.05 andPAF0.1, the C =O stretching vibration peaks red shift toFIGURE 2 (a) FTIR spectra of PFSF, AA, and PAF0.1. (b) The F 1s XPS spectrum of PAF0.1. (c) FTIR spectra of PAF0.1, PAF0.05, andPAA. (d) The TGA and corresponding differential thermal analysis curves of PAF0.1. (e) The DSC curve of PAF0.1, PAF0.05, and PAA. AA,acrylic acid; DSC, differential scanning calorimetry; FTIR, Fourier transform infrared spectroscopy; PAA, polyacrylic acid; PFSF, perfluoro(4‐methyl‐3,6‐dioxaoct‐7‐ene) sulfonyl fluoride; TGA, thermogravimetric analysis; XPS, X‐ray photoelectron spectroscopy.860 | 27693325, 2024, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cnl2.158 by National Institute For, Wiley Online Library on [21/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fcnl2.158&mode=1686 and 1684 cm−1, respectively. The slight red shift ofC =O proves the participation of C =O in the formationof hydrogen bonds. Additionally, the magnitude of redshift and intensity of the characteristic peak increasewith the proportion of PFSF. Furthermore, the stretchingvibration peak of ‐OH in COOH displays a broadeningeffect. The asymmetric and symmetric stretching vibra-tions of ‐SO2‐ in the sulfonyl fluoride group exhibit a redshift in PAFx, occurring at 1449 and 1227 cm−1,respectively. Moreover, the stretching vibration peak ofthe S‐F bond in the sulfonyl fluoride group (located at791 cm−1) also exhibits a red shift phenomenon. Inconclusion, the observed red shift and broadeningphenomena of the characteristic peaks in PAFx areindicative of the presence of hydrogen bonding interac-tions between the sulfonyl fluoride group (‐SO2F) inPFSF and the carboxyl group (‐COOH) in AA.Thermal stability assessment of PAF0.1 was con-ducted using thermogravimetric analysis (TGA). Asdepicted in Figure 2d, PAF0.1 exhibits characteristicdegradation behavior across multiple stages. Beyond160°C, the copolymer undergoes a minor degradationattributed to water loss. Upon surpassing 160°C, thecopolymer initiates its first degradation stage, whereasthe anhydration reaction involving carboxyl groupspersists until temperatures exceed 283°C.[12] The thirdstage of decomposition ranges from 300°C to 363°C,mainly from the decomposition of acrylic anhydride. Thesubsequent degradation stage between 365°C and 460°Carises from the decomposition of the sulfonyl fluoridegroup.[14] With further temperature escalation, PAF0.1copolymer begins to carbonize.[15] Notably, the TGAcurve illustrates the copolymer's retention of thermalstability across a broad temperature range below 160°C, afeature advantageous for alleviating safety concerns inlithium‐ion batteries.The glass transition temperature (Tg) of the polymerswas investigated using differential scanning calorimetry.As depicted in Figure 2e, PAA exhibits a higher Tg valueof 136.5°C compared to PAF0.05 and PAF0.1, with Tgvalues of 124.8°C and 120.5°C, respectively. Thisphenomenon arises due to the exclusive presence ofC =O ∙ ∙ ∙H‐OOC hydrogen bonds between AA mole-cules within the PAA binder, exhibiting a remarkablyhigh bond energy of 20.08 kcal mol−1. However, upon theintroduction of PFSF molecules into the polymer binder,S = O ∙ ∙ ∙H‐OOC hydrogen bonds with a bond energy ofonly 2.39 kcal mol−1 are generated between PFSF and AAmolecules. In PAF0.05 and PAF0.1 binders, because someAA molecules participate in the formation ofS =O ∙ ∙ ∙H‐OOC hydrogen bonds, the proportion ofC =O ∙ ∙ ∙H‐OOC hydrogen bonds between AA and AAmolecules decreases. Substitution of the high bondenergy C =O ∙ ∙ ∙H‐OOC hydrogen bond with the lowerbond energy S =O ∙ ∙ ∙H‐OOC hydrogen bond leads to areduction in the intermolecular force within the PAFxstructure. Notably, as the proportion of PFSF monomersincreases, there is a corresponding increase in thereplacement of C =O ∙ ∙ ∙H‐OOC hydrogen bonds withS =O ∙ ∙ ∙H‐OOC hydrogen bonds, further weakeningthe intermolecular forces within PAFx. Consequently, theTg experiences a gradual decline in PAA, PAF0.05, andPAF0.1 binders due to these structural modifications.In the spectra of pure PAF0.1 and PAF0.1 combinedwith SiOx as a binder (Supporting Information S1:Figure S3), noticeable differences are observed. Specifi-cally, compared to the pure PAF0.1 binder, a distinct ‐COO‐ peak is observed at 1722 cm−1 in the combinedspectrum. This observation serves as evidence of anesterification reaction occurring between the ‐COOHfunctional group of the PAF0.1 binder and the ‐OHgroups of SiOx. Additionally, several spectral changes arenoted: the stretching vibration of ‐OH in ‐COOH, thesymmetrical absorption peak of ‐SO2, and the stretchingvibration peak of the S‐F bond all exhibit broaderprofiles. These broadening effects are indicative of thepresence of hydrogen bonds between the ‐SO2F and ‐COOH functional groups of the binder and the ‐OHgroups of SiOx.Supporting Information S1: Figure S4 shows thescanning electron microscopy (SEM) image and energyspectrum of the CNT‐NH2. As shown in the figure, thediameter of the carbon nanotube is about 50 nm and thelength is several microns. The energy spectrum resultsshow that the nitrogen element on the surface of thecarbon nanotube is evenly distributed in a point‐likemanner, indicating that ‐NH2 is evenly distributed on thesurface of the carbon nanotube. To explore the interac-tion between the polymer binder and the CNT‐NH2conductive agent, XPS tests were performed on theirmixtures. For instance, the N 1s spectrum (SupportingInformation S1: Figure S5) reveals distinct character-istics: solely the C‐N peak at 399.2 eV is discerned forpure CNT‐NH2. However, upon the addition of PAA orPAFx, a prominent O = C‐N covalent bond peak emergesat 398.5 eV, as depicted in Figure 3a–c. Moreover, theintroduction of PFSF monomer causes the area of thecovalent O = C‐N peak to gradually increase. Thisincrease signifies a rising degree of amidation betweenthe polymer binder and the CNT‐NH2 conductive agent,indicating that PFSF can promote the amidation reactionbetween AA and CNT‐NH2. This effect may stem fromthe highly electronegative perfluorinated side chainpresent in the PFSF monomer, which exerts a strongelectron‐withdrawing influence. This renders the ‐COOHgroup in the AA monomer more acidic, thus facilitating| 861 27693325, 2024, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cnl2.158 by National Institute For, Wiley Online Library on [21/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fcnl2.158&mode=its reactivity in undergoing an amidation reaction withthe ‐NH2 groups of carbon nanotubes. Furthermore, thisphenomenon is also evident in the C 1 s spectrum of theirmixture (Supporting Information S1: Figure S6).[16]To delve further into the interaction between thePAF0.1 binder and the CNT‐NH2 conductive agent, FTIRtests were conducted on their mixtures. The results, asdepicted in Figure 3d, reveal distinct peaks in the rangeof 1500–1900 cm−1, attributed to various componentssuch as C = C, COOH, C =O, and N‐H functionalities.[17]These peaks correspond to the carbon skeleton of CNT‐NH2, the carboxyl group of the AA monomer, the amide Iband (C =O), and the amide II band (N‐H), respectively.Upon comparison with the spectrum of pure PAF0.1(Figure 3e), noticeable changes are observed. Specifically,a red shift in the ‐COOH peak indicates the formation ofa hydrogen bond between the ‐COOH group of thePAF0.1 binder and the ‐NH2 group of the CNT‐NH2conductive agent. Additionally, the appearance of theamide I band (C =O) and the amide II band (N‐H)further confirms the occurrence of an amidation reactionbetween the ‐COOH group of the PAF0.1 binder and the ‐NH2 group of the CNT‐NH2 conductive agent.[18] Asdepicted in Figure 3f, distinct peaks corresponding to theasymmetric stretching vibration of ‐SO2‐ and thestretching vibration of S‐F are observed at 1441 and790 cm−1, respectively, within the range of700–1500 cm−1. Compared with the spectrum of purePAF0.1, the red shift of the ‐SO2‐ peak and the S‐F peakindicates the formation of hydrogen bonds between the ‐SO2‐ group of the PAF0.1 binder and the ‐NH2 group ofthe CNT‐NH2.FIGURE 3 The comparison of XPS N 1s spectra of (a) PAA‐CNT‐NH2, (b) PAF0.05‐CNT‐NH2, (c) PFA0.1‐CNT‐NH2. FTIR spectra of (d)PAA‐CNT‐NH2, (e) PAF0.05‐CNT‐NH2, (f) PFA0.1‐CNT‐NH2. CNT‐NH2, amino‐functionalized carbon nanotube; FTIR, Fourier transforminfrared spectroscopy; PAA, polyacrylic acid; XPS, X‐ray photoelectron spectroscopy.862 | 27693325, 2024, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cnl2.158 by National Institute For, Wiley Online Library on [21/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fcnl2.158&mode=2.3 | Mechanical propertiesFigure 4a depicts the load‐displacement curves obtainedfrom nano‐indentation tests on two polymer films, alongwith their respective hardness and elastic modulusvalues.[19] The PAA film demonstrates calculated valuesof 381.1MPa for hardness and 9.34 GPa for elasticmodulus, as depicted in Figure 4b. In contrast, PAF0.1shows lower hardness (316.6MPa) and elastic modulus(8.67 GPa). The diminished elastic modulus and hardnessobserved in PAF0.1 can be attributed to the introductionof flexible PFSF monomer, which makes C =O ∙ ∙ ∙H‐OOC with high bond energy (20.08 kcal mol−1) beingreplaced by S =O ∙ ∙ ∙H‐OOC with low bond energy(2.39 kcal mol−1), reducing the intermolecular force. Thelow Young's modulus indicates that the PAF0.1 binderhas better deformation ability. When SiOx volumeexpands, it can effectively dissipate internal stressthrough the gradient dissociation and reconstruction ofhydrogen bonds between ‐COOH and ‐SO2F functionalgroups.The inclusion of PFSF is expected to have asubstantial impact on the rheological properties ofPAF0.1 polymer molecules, a phenomenon that can beconfirmed through rheometer analysis (Figure 4c). At thesame concentration, both storage modulus (G’) and lossmodulus (G”) of the PAF0.1 solution are higher thanthose of the PAA solution, indicating that the PAF0.1solution has enhanced elastic deformation recovery andviscous deformation dissipation capabilities.[19a] Thismay be attributed to the gradient dissociation andreconstruction of hydrogen bonds between functionalgroups under shear stress. The complex viscosity of thetwo polymer solutions was measured at a concentrationFIGURE 4 (a) The load‐displacement curves of PAA and PAF0.1 films in nano‐indentation test and (b) corresponding hardness andelastic modulus. (c) The rheological properties of 5 wt% PAA and PAF0.1 aqueous solution. (d) Complex viscosities of 5 wt% PAA and PAF0.1aqueous solution. (e, f) 180° peeling curves and average peeling force of SiOx electrodes with PAA‐CB and PAF0.1‐CNT. CB, acetylene black;CNT, carbon nanotube; PAA, polyacrylic acid.| 863 27693325, 2024, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cnl2.158 by National Institute For, Wiley Online Library on [21/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fcnl2.158&mode=of 5 wt% (Figure 4d). The complex viscosity of PAF0.1 ishigher from low frequency (0.1 Hz, 0.1009 Pa s) to highfrequency (10 Hz, 0.3080 Pa s), whereas the values forPAA are lower than 0.0111 Pa s (0.1 Hz) and 0.0129 Pa s(10 Hz).In general, the viscosity of PAF0.1 is ~23 times greaterthan that of PAA, suggesting that PAF0.1 experiencessignificantly higher internal friction. Mainly due to thelonger perfluorinated side chains of PAF0.1, it is morelikely to become entangled when subjected to shearstress, resulting in an increase in viscosity. The adhesioncapability of binders plays a critical role in preventingactive material detachment and maintaining electrodeintegrity. To assess their adhesion ability, typical 180°peel tests were conducted on SiOx electrodes usingdifferent binders. As depicted in Figure 4e,f, the PAF0.1‐CNT electrode (15.31 N) consistently exhibited higherpeel force throughout the entire peel‐off process, nearlynine times that of the PAA‐CB electrode (1.65 N). Thisresult underscores the superior adhesion capability of thePAF0.1 binder over PAA binder. Remarkably, thisexcellent performance is achieved with only 10% contentof the binder PAF0.1, which is significantly less than the20% content of PAA‐CB. The primary reason for thedifference in performance between the PAA‐CB andPAF0.1‐CNT anodes lies in the choice of conductiveagent. The PAA‐CB anode utilizes traditional CB, whichlacks polar groups on its surface and can only establish aconnection with the binder through weak van der Waalsforces. In contrast, the PAF0.1‐CNT anode employs anamino‐functionalized CNT‐NH2 conductive agent, char-acterized by numerous ‐NH2 on its surface. These ‐NH2enable the conductive agent to form connections with thebinder through gradient hydrogen bonds and covalentamide bonds, resulting in the formation of a robust three‐dimensional network structure. This fully proves that thethree‐dimensional network structure mediated by physi-cal and chemical double cross‐linking has uniqueadvantages in improving the electrode connectionstrength and maintaining the integrity of the electrodestructure.2.4 | Electrochemical performances ofdifferent electrodesTo investigate the reduction stability of the adhesive,linear sweep voltammetry testing was initially conductedon the binder film coated on copper foil. The results, asdepicted in Supporting Information S1: Figure S7, showthree distinct reduction peaks at 1.73, 1.47, and 0.78 V,respectively, as the potential sweeps from 3 to 0 V. Thesepeaks correspond to the reduction and decompositionprocesses of the binder, FEC, and EC, respectively.[20]The observation that the binder decomposes preferen-tially over the electrolyte aligns with the theoreticalcalculation results. Compared to the inorganic compo-nents such as carboxylates generated from the decompo-sition of solvents like EC and FEC, the binder'sdecomposition, rich in ‐SO2F groups, leads to theformation of inorganic components such as Li2O, Li2SO3,and LiF. These inorganic components contribute to ahigher mechanical strength of the SEI, aiding inmaintaining the structural integrity of silicon oxideparticles.The electrochemical performance of SiOx‐basedelectrodes using various binders and conductive agentswas systematically evaluated through a series of half‐celltests. To verify the uniform electrochemical behavior ofSiOx particles, a cyclic voltammetry (CV) test wasconducted using PAF as the binder, and the results areillustrated in Figure 5a. In the initial cycle, two distinctreduction peaks are evident at 1.89 V and 1.39 V in thePAF0.1‐CNT anode. These peaks correspond to thereduction and decomposition processes of the PAF0.1binder and the electrolyte on the SiOx surface, respec-tively.[21] This observation aligns well with the theoreti-cal calculation results. Furthermore, a conspicuousreduction peak is observed at 0.15 V, indicative of theformation of a Li‐Si alloy. Moreover, broader peaksaround 0.34 V and 0.53 V were identified as the delithia-tion stages of the Li‐Si alloy. During the subsequent scancycles, it was noted that the reduction peak at 1.39 Vdisappeared. However, the reduction peak at 1.89 Vpersisted and shifted slightly to 1.83 V. This shiftindicates that the continuous reduction and decomposi-tion of the binder led to the formation of a stablestructure on the SiOx surface, known as the SEI. Thisstable SEI structure serves to prevent the continuousdecomposition of the electrolyte, contributing to im-proved stability and performance of the anode. Moreover,the intensity of three main redox peaks mentioned aboveprogressively increased with cycling, indicating theinfiltration of electrolyte into the electrode and theelectrochemical activation of SiOx. These findings pro-vide valuable insights into the electrochemical behaviorand activation mechanisms of SiOx‐based electrodes,highlighting the role of PAF as a binder in facilitatingstable electrochemical processes.The initial charge‐discharge curve of the SiOx anodewith PAA‐CB, PAA‐CNT, PAF0.05‐CNT, or PAF0.1‐CNTat a rate of 0.1 C is shown in Figure 5b. Notably, animportant observation is that the ICE of the PAF systemanode (72.66% and 74.11%) exceeds that of the PAAsystem anode (70.47% and 70.55%). Furthermore, it'snoted that the higher the PFSF content is, the higher the864 | 27693325, 2024, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cnl2.158 by National Institute For, Wiley Online Library on [21/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fcnl2.158&mode=ICW will be. On the one hand, this is due to the increasein PFSF monomer and the gradual increase in the degreeof amidation reaction between the binder and carbonnanotubes, forming a tighter three‐dimensional conduc-tive network, which can provide efficient conductivepathways and maintain a more complete electrodestructure. On the other hand, it may be that the PAF0.1binder has a lower LUMO energy level and decomposespreferentially than the electrolyte, promoting the forma-tion of a stable SEI.Figure 5c illustrates the constant current cyclingperformance of SiOx anodes utilizing PAA‐CB, PAA‐CNT, PAF0.05‐CNT, or PAF0.1‐CNT at a rate of 0.5 C. Theinitial discharge capacities for PAA‐CB, PAA‐CNT,PAF0.05‐CNT, and PAF0.1‐CNT anodes are recorded as1760.8, 1718.4, 1888.2, and 1830.5 mAh g−1, respectively.Throughout subsequent cycles, PAA‐CB and PAA‐CNTanodes exhibit rapid capacity decay, with PAA‐CNTdemonstrating a faster decline. In contrast, the capacitydecay of PAF0.05‐CNT and PAF0.1‐CNT anodes isrelatively slower. In the first 250 cycles, the capacity ofthe PAA‐CB anode dropped rapidly to below 500mAhg−1 (capacity retention rate 27.1%), and it remainedbasically stable until 300 cycles. After 500 cycles, it onlymaintained a capacity of 308.8 mAh g−1 (capacity reten-tion rate 17.5%) and the average Coulombic efficiencyonly 99.66%. The rapid capacity loss observed in thePAA‐CB anode can be attributed to the partial breakageof covalent bonds between hydroxyl and carboxyl groupsfollowing the volume expansion of SiOx. Moreover, thereis only a weak van der Waals force between the CBconductive agent, PAA binder, and SiOx particles, whichcannot form an effective conductive path. The capacity ofthe PAA‐CNT anode experienced a rapid decline,dropping below 500mAh g−1 (capacity retention rate28.3%) after only 20 cycles. Eventually, it stabilized at acapacity of 141.8 mAh g−1 (capacity retention rate 8.2%)after 500 cycles, with an average Coulombic efficiency asFIGURE 5 (a) CV profiles of SiOx@PAF0.1‐CNT anode. (b) Galvanostatic charge–discharge profiles of different anodes at the first cycle.(c) Cycling performances of different anodes at 0.5 C. (d) Rate performances of different anodes. (e) Cycling performances of differentanodes at 1 C. CB, acetylene black; CNT, carbon nanotube; CV, cyclic voltammetry; PAA, polyacrylic acid.| 865 27693325, 2024, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cnl2.158 by National Institute For, Wiley Online Library on [21/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fcnl2.158&mode=low as 99.54%. This decline can be attributed to severalfactors. First, the low binder content in the PAA‐CNTnegative electrode hinders its ability to effectivelyconnect with SiOx particles. This lack of connectivityresults in reduced capacity and performance over cycles.Additionally, the presence of only strong covalent bondsbetween the binder and the carbon nanotube conductiveagent is insufficient to withstand the volume expansionof SiOx. Without an effective mechanism to buffer stressduring expansion, the covalent bonds eventually break,further contributing to the decline in capacity andefficiency of the anode. Conversely, the incorporationof sulfonyl fluoride groups mitigated this issue byenhancing hydrogen bonds with carboxyl and aminogroups. As a result, the reversible capacity of the PAF0.05‐CNT anode improved significantly to 580.1 mAh g−1(capacity retention rate 30.7%), with an average Coulom-bic efficiency of 99.74% after 500 cycles. Moreover, as thecontent of sulfonyl fluoride groups increased, thereversible capacity also increased. Notably, the PAF0.1‐CNT anode exhibited the highest reversible capacity of627.3 mAh g−1 (capacity retention rate 34.3%), with anaverage Coulombic efficiency value of 99.76%. The aboveresults fully prove that at lower binder and conductiveagent contents, the introduction of PFSF monomer canenable the binder to form a covalent–noncovalent three‐dimensional conductive network through gradienthydrogen bonds and covalent amide bonds, which iseffective Buffers SiOx volume expansion and improvescycle stability.Figure 5d illustrates the rate performance of variousanodes. The PAF0.1‐CNT anode demonstrates capacitiesof 1477.4, 1343.3, 1214.6, 1003.2, and 797.6 mAh g−1 atrates of 0.1, 0.5, 1, 2, and 3 C, respectively. Additionally,upon reverting the cycling rate to 0.1 C, the PAF0.1‐CNTanode maintains a substantial capacity of 1392.8 mAhg−1. In contrast, the PAA‐CNT anode exhibited capacitiesof 282.5, 135.8, and 82.4 mAh g−1 at rates of 1, 2, and 5 C,respectively. Upon returning to 0.1 C, it retained acapacity of 708.4 mAh g−1, showing a rapid decreasingtrend. The huge contrast in rate performance betweenPAA‐CNT anode and PAF0.1‐CNT anode under highcurrent density fully proves that the huge volumeexpansion of SiOx cannot be fully accommodatedthrough the irreversible covalent three‐dimensional net-work alone. The stable SEI and covalent‐noncovalentsynergy formed after the introduction of PFSF monomercan better maintain the three‐dimensional conductivenetwork structure and improve the rate performanceof SiOx.To further verify the long‐term cycle stability ofdifferent binders, a constant current cycle at a rate of 1 Cwas performed in Figure 5e. All cells were first cycled at0.1 C for three cycles to activate and subsequently at 1 C.After activation, the reversible discharge capacities ofSiOx electrodes with PAA‐CB, PAA‐CNT, PAF0.05‐CNT,and PAF0.1‐CNT were 1156.6, 843.5, 1034.8, and1104.5 mAh g−1, respectively. After 800 cycles, thePAF0.1‐CNT anode showed a reversible capacity of502.9 mAh g−1, which was much higher than PAA‐CB(280.7 mAh g−1) and PAA‐CNT (60.7 mAh g−1). ThePAF0.1‐CNT anode showed an excellent capacity reten-tion of 84.9% over the first 300 cycles, retaining 45.5% ofthe capacity after 800 cycles relative to activation. Incontrast, the PAA‐CB anode only had a capacityretention of 24.3% after 800 cycles. What's more, thePAA‐CNT anode experienced severe capacity loss in thefirst 30 cycles, maintaining a capacity of 280.1 mAh g−1 atthe 30th cycle and a low capacity retention rate of 7.2%after 800 cycles. To improve the performance of the SiOxanode, we assembled a complete battery by pairingdifferent anodes with LiNi0.6Co0.2Mn0.2O2 (NCM) cath-odes. Supporting Information S1: Figure S8 shows that at0.5 C, the PAF0.1‐CNT|NCM full cell exhibits significantlymore reversibility and lower capacity attenuation (0.38%/cycle) than PAA‐CB | NCM over 50 cycles, and morestable cycle performance (1C = 200mAh g−1). This resultproves that the physicochemical dual crosslinkingconductive network of PAFx‐CNT can effectively bufferenormous volume changes of SiOx by the gradienthydrogen bond and covalent bond among ‐COOH,‐SO2F, and ‐NH2.High‐resolution XPS analysis was employed toexamine the interfacial stability of the SEI on the SiOxelectrode surface following 50 cycles. Within the C 1sspectrum, discernible peaks corresponding to C−H/C−C, C−O, C =O, and Li2CO3/ROCO2Li are identifiedin the PAA‐CB electrode at 284.8, 286.1, 287.1, and290.2 eV, respectively (Figure 6a,b). These organicconstituents within the SEI are indicative of electrolytedecomposition. Conversely, a reduction in peak intensityfor each C peak was observed in the PAF0.1‐CNTelectrode, suggesting that the surface‐stable SEI effec-tively mitigates electrolyte decomposition. In the O 1sspectrum (Figure 6c,d), heightened intensities of peaks at531.7 eV (Li2CO3) and 529.7 eV (Li2O) are observed onthe PAF0.1‐CNT electrode. It is worth noting that thepeak of Li2CO3 is shifted relative to the correspondingpeak in PAA‐CB, which may be due to the introductionof Li2SO3 into the SEI by the modified additives. In the F1s spectrum (Figure 6e,f), two distinct peaks emerged:LixPOyFz (687.2 eV) and LiF (684.6 eV), indicative of FECand LiPF6 decomposition. However, in the F 1s XPSspectrum of the PAF0.1‐CNT anode, the LiF peakintensity is higher, which is beneficial to improving thestability of SEI, and an obvious RSO2F peak appears at866 | 27693325, 2024, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cnl2.158 by National Institute For, Wiley Online Library on [21/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fcnl2.158&mode=688.7 eV, which originates from incomplete decomposi-tion of sulfonyl fluoride binder. As shown in SupportingInformation S1: Figure S9, compared with PAA‐CB, theLiF content on the surface of PAF0.1‐CNT anodeincreased by 4.5 times, while the Li2O content increasedby 1.6 times, and the Li2SO3/Li2CO3 content onlyincreased by 10%. In the C 1s spectrum, compared withPAA‐CB, the peak intensity of Li2CO3/ROCO2Li on thesurface of PAF0.1‐CNT anode is lower, indicating that thecontent of Li2CO3 on the surface of PAF0.1‐CNT anode islower. Therefore, the increase in Li2SO3/Li2CO3 contentmainly comes from Li2SO3 produced by the decomposi-tion of ‐SO2F group. The generation of Li2SO3 with highion conductivity makes it easier for lithium ions to passthrough the SEI and participate in the lithium deinter-calation reaction of SiOx, which is beneficial to reducingthe RSEI and Rct of the PAF0.1‐CNT negative electrode.The above results fully prove that the ‐SO2F group in thePAF0.1 binder with lower LUMO preferentially partici-pates in the formation of SEI, promoting the generationof more nonpolar components such as Li2O, Li2SO3/Li2CO3 and LiF.To fully understand the composition and structure ofSEI, the electrodes of different systems were character-ized by XPS sputtering after 50 cycles. As shown inSupporting Information S1: Figure S10, with the exten-sion of sputtering time, the content of LiF graduallyincreased, but the content of Li2CO3 and Li2O did notincrease significantly, and the content of C‐O and otherorganic components gradually increased. Even if thesputtering time reached 120 s, some LixPFyOz stillexisted. This indicates that the PAA‐CB system cannotinhibit the penetration of the electrolyte into theelectrode, which leads to the decomposition of theFIGURE 6 (a, b) C 1s, (c, d) O 1s, and (e, f) F 1s XPS spectra of various anodes after 50 cycles. CB, acetylene black; CNT, carbonnanotube; PAA, polyacrylic acid; XPS, X‐ray photoelectron spectroscopy.| 867 27693325, 2024, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cnl2.158 by National Institute For, Wiley Online Library on [21/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fcnl2.158&mode=infiltrated electrolyte into the inorganic componentssuch as LixPFyOz and ROCO2Li in the electrode. Thiskind of organic‐inorganic co‐existing SEI has lowmechanical strength, which is not conducive to main-taining the stability of electrode structure. In contrast(Supporting Information S1: Figure S11), with theextension of sputtering time in PAF0.1‐CNT system, thecontents of inorganic components such as LiF, Li2CO3/Li2SO3, and Li2O gradually increased, while the contentsof organic components such as LixPFyOz and C‐Ogradually decreased. When the sputtering time reached120 s, the contents of inorganic components such asLixPFyOz and C‐O gradually decreased. The LixPFyOzsignal is almost completely gone, only the LiF signal. Theresults show that the PAF0.1‐CNT system can effectivelyinhibit the electrolyte from penetrating into the electrodeand avoid its decomposition into inorganic componentssuch as LixPFyOz and ROCO2Li. In addition, thepreferential decomposition of the binder in the electrodealso promotes the formation of more inorganic compo-nents such as Li2CO3/Li2SO3 and Li2O. This kind ofinternally rich inorganic SEI has high mechanicalstrength and is conducive to maintaining the stabilityof the electrode structure.2.5 | Elastic properties of differentelectrodesAs previously mentioned, the PAFx‐CNT double cross‐linked binder plays a pivotal role in stabilizing theelectrode structure and mitigating volume changes. Todelve deeper into the impact of physical and chemicaldouble cross‐linking on the macroscopic and electrodesurface state postcycling, SEM was employed to comparesurface morphology pre and post cycling. Figure 7illustrates that the surfaces of both the PAA‐CB andPAA‐CNT anodes appeared rougher precycling com-pared to other variants. After 50 cycles, extensive crackswere observed on the PAA‐CB and PAA‐CNT anodes,especially the latter. This phenomenon may arise frompoor compatibility between PAA fragments and SiOxparticles. Moreover, the micron‐sized SiOx particles inPAA‐CNT are coated with a covalent network, resultingin an uneven surface prone to covalent bond breakagedue to insufficient elasticity. In contrast, micrometer‐sized pores visible precycling on the electrodes vanishedon the PAFx‐CNT anode post‐cycling. The healing ofcracks on PAFx‐CNT anodes can be attributed to theirintroduction of more elastic noncovalent networks. Theexcellent elastic properties of this binder contribute tobetter cycling performance of PAFx‐CNT anodes. FurtherSEM analysis of cross‐sectional morphology revealed thatpre‐cycling, the pristine PAA‐CB and PAA‐CNT anodeshad thicknesses of 21.06 and 22.12 μm, respectively.Postcycling, their thicknesses significantly increased to33.65 and 36.06 μm, respectively, with expansion rates of59.8% and 63.0% (Figure 7m,n). In contrast, the PAF0.05‐CNT and PAF0.1‐CNT anodes only showed modestincreases to 32.21 and 31.25 μm, with thickness increaserates of 48.9% and 39.8%, respectively (Figure 7o,p).Compared with the PAA‐CB anode, the PAA‐CNT anodehas more and deeper surface cracks and a higherexpansion rate, indicating that the covalent three‐dimensional conductive network structure alone cannotsuppress volume expansion and maintain the integrity ofthe electrode structure. The surfaces of PAF0.05‐CNT andPAF0.1‐CNT negative electrodes are smoother and havelower expansion rates. Mainly because after the intro-duction of PFSF monomer, the flexibility of the polymerbinder increases. And the results show that the physico-chemical double cross‐linking mediated covalent andnoncovalent three‐dimensional network can maintainthe integrity of electrode structure and promote theformation of stable SEI.To explore the relationship between the excellentelectrochemical performance of the PAF0.1‐CNT anodeand its ionic conductivity, the Li+ diffusion coefficient(DLi+) of SiOx anode was meticulously examined usingCV at varying scan rates. Supporting Information S1:Figure S12 showcases CV curves for the PAA‐CB, PAA‐CNT, PAF0.05‐CNT, and PAF0.01‐CNT anodes across scanrates ranging from 0.2 to 1.2 mV s−1. The PAA‐CNTanode exhibits an approximately fourfold increase inpeak current intensity compared to the PAA‐CB anode,demonstrating a substantial enhancement in electronicconductivity facilitated by the three‐dimensional conduc-tive network established by the CNT‐NH2 conductiveagent within the electrode structure. Furthermore, theintroduction of the PFSF monomer leads to a progressiveincrease in peak current intensity for both the PAF0.05‐CNT and PAF0.1‐CNT anodes, providing evidence thatthe incorporation of PFSF monomer contributes to theimprovement of ionic conductivity. Specifically, inSupporting Information S1: Figure S12e,f, the oxidationpeak current (IA) shows a positive correlation with thesquare root of the scan rate (ν1/2), whereas the reductionpeak current (IC) exhibits a negative correlation with ν1/2.The PAA‐CB anode exhibited the lowest Li+ diffusioncoefficient (DLi+) during both lithiation and delithiationprocesses. However, the introduction of carbon nano-tubes significantly increased the DLi+ of the PAA‐CNTanode by one to two orders of magnitude, underscoringthe profound impact of this three‐dimensional conduc-tive network on battery performance enhancement. ThePAA‐CB anode exhibited the lowest Li+ diffusion868 | 27693325, 2024, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cnl2.158 by National Institute For, Wiley Online Library on [21/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fcnl2.158&mode=coefficient (DLi+) during both lithiation and delithiationprocesses. However, the introduction of CNT‐NH2significantly increased the DLi+ of the PAA‐CNT anodeby one to two orders of magnitude, underscoring theprofound impact of this three‐dimensional conductivenetwork on battery performance enhancement. More-over, the incorporation of the PFSF monomer into thecopolymer binder further improved the DLi+ of the PAFx‐CNT anode. This improvement can be attributed to thepresence of the PFSF monomer, which is rich inperfluoropolyether side chains capable of participatingin Li+ coordination and transport through polar C‐O‐Cbonds and ‐SO2F functional groups. This facilitates therapid diffusion of Li+, contributing to the enhanced Li+conductivity observed in the PAFx‐CNT anode.[22]Furthermore, to account for the variation in Li+diffusion kinetics during different charge and dis-charge stages, real‐time analysis of Li+ diffusionFIGURE 7 (a–h) SEM images of different anodes before cycling and after 50 cycles. (i–p) Cross‐sectional SEM images of different anodesbefore cycling and after 50 cycles. CB, acetylene black; CNT, carbon nanotube; PAA, polyacrylic acid; SEM, scanning electron microscopy.| 869 27693325, 2024, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cnl2.158 by National Institute For, Wiley Online Library on [21/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fcnl2.158&mode=kinetics during cycling was conducted via galvano-static intermittent titration technique as illustrated inSupporting Information S1: Figure S14a,b. Initially,all four anodes showed comparable Li+ diffusioncoefficients (DLi+) due to insufficient activation of theSiOx anode.[23] However, after five cycles, the DLi+ ofPAA system anodes are significantly lower than thatof PAFx anodes and its DLi+ decreased compared withbefore the cycle. This is mainly due to the fact that thecovalent bond of the SiOx anodes with PAA as thebinder cannot withstand the huge volume expansion,resulting in the covalent bond breaking, and even-tually the particle crushing and subsequent loss ofelectrical contact. When PFSF monomer is intro-duced, the DLi+ of the PAFx‐CNT anode is two to threetimes that of the PAA‐CB and PAA‐CNT anodes. Thisfurther demonstrates the importance of the physicaland chemical double‐cross‐linked three‐dimensionalconductive network of the PAFx‐CNT anode forbuffering the volume expansion of SiOx (maintainingthe integrity of the electrode structure) and theinvolvement of polar C‐O‐C bonds and ‐SO2F func-tional groups in Li+ coordination and facilitating Li+transport.To further investigate the impact of the PAFx binderon impedance and Li+ diffusion processes in SiOx anodesbefore and after cycling, the SiOx anode underwentelectrochemical impedance spectroscopy (EIS) analysis,measuring resistance before and after cycling. Theequivalent circuit model depicted in Supporting Infor-mation S1: Figure S15 was used, with resulting resist-ances detailed in Supporting Information S1: Table S1.Initially, EIS curves (Supporting Information S1:Figure S16a) for all SiOx anodes displayed a semicircleand a linear segment, representing charge transferresistance (Rct) and ion diffusion resistance (Zw),respectively.[24] It is notable that the semicircle diameterof the PAA‐CB anode exhibits a significant increase, withthe charge transfer resistance (Rct) measuring 178.60Ω,surpassing that of other anodes. Conversely, the Rct of thePAF0.1‐CNT anode is notably reduced, measuring onlyone‐fourth of that observed in the PAA‐CB anode.Postcycling (50 cycles), additional semicircles (RSEI)emerged in the high‐frequency region of EIS curves(Supporting Information S1: Figure S16b) for all SiOxanodes. The PAF0.1‐CNT anode stood out with the lowestRSEI (13.11Ω) and Rct (4.67Ω), showcasing superiorresistance characteristics both pre‐ and postcycling. Thisenhanced performance in the PAF0.1‐CNT anode can beattributed to the unique physicochemical double cross‐linked structure derived from the PAF0.1‐CNT system,along with the rich‐inorganic SEI induced by the sulfonylfluoride group.3 | CONCLUSIONThe gradient hydrogen and covalent amide bonds formedby the ‐SO2F and ‐COOH of PAF0.1 with the ‐NH2 ofCNT‐NH2 participate in the formation of physical andchemical double‐crosslinked three‐dimensional conduc-tive networks, which give the SiOx anode significantlyenhanced performance. As shown in Figure 8, when thevolume of SiOx expands, the stress can be effectivelydissipated through the stepwise dissociation of gradienthydrogen bonds with bond energies in a wide range(2.39–20.08 kcal mol−1), maintaining the stability of theelectrode structure and ensuring improve the long‐termcycle stability of SiOx anode. Simultaneously, the bindermolecules possess a lower LUMO energy level comparedto all solvent molecules. This characteristic predisposesthem to preferentially undergo reduction on the anodesurface, leading to the formation of inorganic compo-nents rich in Li2O, Li2SO3/Li2CO3 and LiF. Thesecomponents aid in reducing SEI impedance and enhan-cing the dynamics of Li+ diffusion.4 | EXPERIMENTAL SECTION4.1 | Synthesis of PAF0.1 copolymerbinderOne gram of AA and either 0.05 g or 0.1 g of PFSF wereseparately placed into a three‐necked flask. Subse-quently, an appropriate amount of deionized water wasadded to dissolve the monomers to a mass fraction of5 wt%. Nitrogen gas was then introduced into the three‐necked flask for 30min to eliminate oxygen. Finally,5 mg of ammonium persulfate was added as a initiatorfor free radical polymerization. Under sealed conditionswith nitrogen gas, the mixture was heated using an oilbath at 50°C for 8 h. Upon completion of the reaction, thesolution was transferred to anhydrous ether, centrifuged,and subjected to repeated washing to obtain a colorlessand transparent precipitate. This precipitate was thendried in a vacuum drying oven at 50°C for 24 h to obtainthe PAFx copolymer. The binders were labeled as PAF0.05and PAF0.1, respectively, based on the mass ratio of PFSFto AA.As shown in Supporting Information S1: Figure S17,the prepared PAF0.1 copolymer has a precisely controlledmolecular weight. The Mn and Mw of PAF0.1 copolymerare 352,469 and 618,204 gmol−1, and the relatively lowpolydispersability index is 1.75. In addition, 1H and19FNMR spectra also supported the successful synthesisof PAF0.1 copolymer (Supporting Information S1:Figures S18 and S19). In summary, free radical870 | 27693325, 2024, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cnl2.158 by National Institute For, Wiley Online Library on [21/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fcnl2.158&mode=polymerization can precisely control the molecularstructure of the polymer.The materials characterizations, mechanical analyses,and electrochemical measurements were conductedfollowing the methodology outlined in our priorwork.[3b,21b]AUTHOR CONTRIBUTIONSGen Chen and Gang Wu conceived the concept. GangWu and Yuanhang Gao designed and carried out theexperiments. Yuanhang Gao conducted the theoreticalcalculations. Zheng Weng and Zhicheng Zheng assistedwith carrying out the experiments and analyzing thedata. Wenqiang Fan conducted verification and investi-gation. Gang Wu and Yuanhang Gao wrote the manu-script. Ning Zhang and Renzhi Ma reviewed and revisedthe manuscript. Gen Chen, Xiaohe Liu and Anqiang Pansupervised the whole project. All authors contributed tothe discussion of the manuscript.ACKNOWLEDGMENTSGang Wu and Yuanhang Gao contributed equally to thiswork. The authors acknowledge the financial support byNational Natural Science Foundation of China(22379166), Natural Science Foundation for DistinguishedYoung Scholars of Hunan Province (2022JJ10089), KeyResearch and Development Program of Hunan Province(2023GK2015), and Central South University Innovation‐Driven Research Programme (2023CXQD034). This workwas supported in part by the High‐Performance Comput-ing Center of Central South University.CONFLICT OF INTEREST STATEMENTThe authors declare no conflict of interests.REFERENCES[1] a) M. Armand, J. M. Tarascon, Nature 2008, 451, 652; b)J. W. Choi, D. Aurbach, Nat. Rev. Mater. 2016, 1(4), 16013; c)B. Scrosati, J. Hassoun, Y. K. Sun, Energy Environ. Sci. 2011,4, 3287; d) B. Zhang, Y. L. Dong, J. R. Han, Y. J. Zhen,C. G. Hu, D. Liu, Adv. Mater. 2023, 35, 2301320; e) Y. Gao,G. Wu, W. Fang, Z. Qin, T. Zhang, J. Yan, Y. Zhong,N. Zhang, G. Chen, Angew. Chem. Int. Ed. 2024, 63,e202403668.[2] a) S. Choi, T. Kwon, A. Coskun, J. W. Choi, Science 2017,357, 279; b) J. Liu, M. Ben, A. Liu, J. Liu, S. Wang,J. Zhang, Chin. Chem. Lett. 2021, 32, 2914; c) Y. Yan,X. Zhao, H. Dou, J. Wei, W. Zhao, Z. Sun, X. Yang, Chin.Chem. Lett. 2021, 32, 910.[3] a) S. Wu, Y. Yang, C. Liu, T. Liu, Y. Zhang, B. Zhang, D. Luo,F. Pan, Z. Lin, ACS Energy Lett. 2021, 6, 290; b) Z. Weng,S. Di, L. Chen, G. Wu, Y. Zhang, C. Jia, N. Zhang, X. Liu,G. Chen, ACS Appl. Mater. Interfaces 2022, 14, 42494; c)G. Wu, Z. Weng, J. Li, Z. Zheng, Z. Wen, W. Fang, Y. Zhang,FIGURE 8 The mechanism of interaction between PAF0.1 binder and SiOx anode. 3D, three‐dimensional; SEI, solid electrolyteinterphase.| 871 27693325, 2024, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cnl2.158 by National Institute For, Wiley Online Library on [21/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fcnl2.158&mode=N. Zhang, G. Chen, X. Liu, ACS Appl. Mater. Interfaces 2023,15, 34852.[4] Y. X. Wang, X. F. Yang, Y. Yuan, Z. Wang, H. Z. Zhang,X. F. Li, Adv. Funct. Mater. 2023, 33, 2301716.[5] a) B. Y. Jin, D. Y. Wang, J. Zhu, H. Y. Guo, Y. Hou, X. Gao,J. G. Lu, X. L. Zhan, X. J. He, Q. H. Zhang, Adv. Funct. Mater.2021, 31, 2104433; b) Y. K. Jeong, J. W. Choi, ACS Nano 2019,13, 8364; c) W. W. Zeng, L. Wang, X. Peng, T. F. Liu,Y. Y. Jiang, F. Qin, L. Hu, P. K. Chu, K. F. Huo, Y. H. Zhou,Adv. Energy Mater. 2018, 8, 1702314.[6] a) X. X. Jiao, J. Q. Yin, X. Y. Xu, J. L. Wang, Y. Y. Liu,S. Z. Xiong, Q. L. Zhang, J. X. Song, Adv. Funct. Mater. 2021,31, 2005699; b) Y. Wang, H. Xu, X. Chen, H. Jin, J. Wang,Energy Storage Mater 2021, 38, 121.[7] J. Tang, J. Zhou, X. Duan, Y. Yang, X. Dai, F. Wu, J. EnergyChem. 2023, 80, 23.[8] a) S. H. Di, D. X. Zhang, Z. Weng, L. Chen, Y. Zhang,N. Zhang, R. Z. Ma, G. Chen, X. H. Liu, Macromol. Chem.Phys. 2022, 223, 2200068; b) P. Li, G. Chen, N. Zhang, R. Ma,X. Liu, Energy & Environmental Materials 2021, 4, 72; c)H. A. Lee, M. Shin, J. Kim, J. W. Choi, H. Lee, Adv. Mater.2021, 33, 2007460.[9] a) B. Zhang, Y. Dong, J. Han, Y. Zhen, C. Hu, D. Liu, Adv.Mater. 2023, 35, 2301320; b) L. Hu, X. Zhang, P. Zhao,H. Fan, Z. Zhang, J. Deng, G. Ungar, J. Song, Adv. Mater.2021, 33, 2104416.[10] a) D.‐Y. Han, I. K. Han, H. Y. Jang, S. Kim, J. Y. Kwon,J. Park, S. Back, S. Park, J. Ryu, Energy Stor. Mater. 2024, 65,103176; b) Y. Wang, X. Yang, Y. Yuan, Z. Wang, H. Zhang,X. Li, Adv. Funct. Mater. 2023, 33, 2301716.[11] K. Varaprasad, N. N. Reddy, S. Ravindra, K. Vimala,K. M. Raju, Int. J. Polym. Mater. 2011, 60, 490.[12] C. Soykan, R. Coşkun, S. Kirbağ, E. Şahin, J. Macromol. Sci. A2007, 44, 31.[13] a) Y. Shen, J. Xi, X. Qiu, W. Zhu, Electrochim. Acta 2007, 52,6956; b) I. Clara, R. Lavanya, N. Natchimuthu, J. Macromol.Sci. A 2016, 53, 492.[14] K. P. Boroujeni, Z. Tohidiyan, A. Fadavi, M. M. Eskandari,H. A. Shahsanaei, Chem. Select 2019, 4, 7734.[15] a) O. Sel, A. Soulès, B. Améduri, B. Boutevin, C. Laberty‐Robert, G. Gebel, C. Sanchez, Adv. Funct. Mater. 2010, 20,1090; b) M. Colpaert, M. Zaton, V. Ladmiral, D. Jones,J. Rozière, B. Ameduri, Polym. Chem. 2019, 10, 2176; c)B. Grabowska, S. Żymankowska‐Kumon, S. Cukrowicz,K. Kaczmarska, A. Bobrowski, B. Tyliszczak, J. Therm.Anal. Calorim. 2019, 138, 4427.[16] J. Xiong, J. Tao, K. Guo, C. Jiao, D. Zhang, H. Lin, Y. Chen,Fibers Polym. 2015, 16, 1512.[17] M. Mokhtarifar, H. Arab, M. Maghrebi, M. Baniadam, Appl.Phys. A Mater. Sci. Process. 2017, 124, 21.[18] S. Çavuş, G. Gürdaǧ, Ind. Eng. Chem. Res. 2009, 48, 2652.[19] a) S. Pal, R. Mondal, S. Guha, U. Chatterjee, S. K. Jewrajka,Polymer 2019, 180, 121680; b) E. Su, O. Okay, React. Funct.Polym. 2018, 123, 70.[20] P. Wang, H. Zhang, X. Nie, T. Xu, S. Liao, Nat. Commun.2022, 13, 3370.[21] a) H. Kim, K. Kim, J. Ryu, S. Ki, D. Sohn, J. Chae, J. Chang,ACS Appl. Mater. Interfaces 2022, 14, 12168; b) Z. Weng,G. Wu, J. Li, Y. Zhang, R. Zhang, N. Zhang, X. Liu, C. Jia,G. Chen, Small Sci. 2024, 4, 2300133.[22] C. Sun, J. Liu, Y. Gong, D. P. Wilkinson, J. Zhang, NanoEnergy 2017, 33, 363.[23] X. Wan, C. Kang, T. Mu, J. Zhu, P. Zuo, C. Du, G. Yin, ACSEnergy Lett. 2022, 7, 3572.[24] a) Z. H. Li, Y. P. Zhang, T. F. Liu, X. H. Gao, S. Y. Li, M. Ling,C. D. Liang, J. C. Zheng, Z. Lin, Adv. Energy Mater. 2020, 10,1903110; b) H. Liu, C. Li, H. P. Zhang, L. J. Fu, Y. P. Wu,H. Q. Wu, J. Power Sources 2006, 159, 717.SUPPORTING INFORMATIONAdditional supporting information can be found onlinein the Supporting Information section at the end of thisarticle.How to cite this article: G. Wu, Y. Gao, Z. Weng,Z. Zheng, W. Fan, A. Pan, N. Zhang, X. Liu, R. Ma,G. Chen, Carbon Neutralization 2024;3:857–872.https://doi.org/10.1002/cnl2.158872 | 27693325, 2024, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/cnl2.158 by National Institute For, Wiley Online Library on [21/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://doi.org/10.1002/cnl2.158https://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fcnl2.158&mode= Binder-induced inorganic-rich solid electrolyte interphase and physicochemical dual cross-linked network for high-performance SiOx anode 1 INTRODUCTION 2 RESULTS AND DISCUSSION 2.1 Theoretical calculation 2.2 Synthesis and characterization of PAF binder 2.3 Mechanical properties 2.4 Electrochemical performances of different electrodes 2.5 Elastic properties of different electrodes 3 CONCLUSION 4 EXPERIMENTAL SECTION 4.1 Synthesis of PAF0.1 copolymer binder AUTHOR CONTRIBUTIONS ACKNOWLEDGMENTS CONFLICT OF INTEREST STATEMENT REFERENCES SUPPORTING INFORMATION