Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Nanoscale insights into the structure of solution-processed graphene by x-ray scattering

Nanoscale insights into the structure of solution-processed graphene by x-ray scattering Chemicalexfoliationisanattractiveapproachforthesynthesisofgrapheneduetoitslowcostand simplicity.However,challengesstillremaininthecharacterizationofsolution-processedgraphene, inparticularwithatomicresolution.Throughthisworkwedemonstratethex-raypairdistribution functionasanovelapproachtostudysolution-processedgrapheneorother2Dmaterialswith atomicresolution,directlyinsolution,producedbyliquid-phaseandelectrochemicalexfoliations. Theresultsshowthedisappearanceoflong-rangeatomiccorrelations,inbothcases,confirming theproductionofsingleandfew-layergraphene.Inaddition,aconsiderableringdistortionhas beenobservedascomparedtographite,irrespectiveofthesolventused:thenormalsurfaceangle tothesheetofthepowdersampleshouldbelessthan6 ,compatiblewithripplesformation observedinsuspendedgraphene.Weattributethiseffecttotheinteractionofsolventmolecules withthegraphenenanosheets. 1.Introduction the intercalated ions, the graphite electrode works as an anode or cathode, hosting oxidation or reduction One of the most attractive and industrially scal- reactions,respectively[12–14].CathodicECgivesrise able methods for graphene production is given to defect-free graphene, but exfoliation does require by chemical exfoliation of graphite [1–4]. This several hours [15, 16], while anodic EC is quick but approach gives rise to graphene dispersions, which givesrisetoslightlyoxidisedgraphene[14]. can be further processed with simple and low-cost Despite the use of simple methods to pro- methods such as drop casting and inkjet printing duce solution-processed graphene, its character- [5–8]. The most used chemical exfoliation isation is very challenging because the nanosheets approaches are liquid phase exfoliation (LPE) and come in different sizes, thicknesses and chemical electrochemical exfoliation (EC). The LPE method functionalisation [10]. In addition, the lack of met- relies on the use of ultrasound and/or shear force to rology standards strongly limits the commercial- exfoliate bulk graphite into graphene suspended in isation of graphene-based products [10, 17, 18]. a suitable solvent, such as N-Methyl-2-pyrrolidone Only recently, the community has defined some (NMP), N,N-dimethylformamide (DMF), and guidelinesforthecharacterisationofgraphene-based Cyrene (Cy) [9–11]. The EC process is based on materials [19], which led to the development of the expanding the graphite layers following the inter- first ISO/IEC standard (ISO/TS 21356-1:2021) for calation of ions and small molecules driven by an measuring the structural properties of graphene external electric field. According to the charge of [20]. In particular, information on the atomic ©2022TheAuthor(s). PublishedbyIOPPublishingLtd 2DMater.10(2023)015006 ZYanetal Figure1.Schematicillustrationoftheexfoliationprocess:(A)sonication-aidedLPEprocessfollowedbycentrifugation;(B)EC processusinggraphiteascathodicelectrode. structureofsolution-processedgrapheneiscurrently 2.Resultsanddiscussion providedbytransmissionelectronmicroscopy,which requires specific sample preparation, and it is time- Solution-processed graphene was produced by using consuming,soitcanonlybeperformedonaselected cathodic EC and LPE methods, as outlined in number of nanosheets. The electron beam can also figures 1(A) and (B), (and supplementary figure damageorchangethestructureofthenanosheets. S1(a)) following the approaches reported previously Synchrotron x-ray based characterisation tech- [12, 29–31]. The solution process graphene in either niques are increasingly employed in material science NMP and Cy are named as Gr LPE (NMP) Gr LPE [21–24]becauseoftheirhighphotonenergy(shorter (Cy), Gr EC (NMP) and Gr EC (Cy), respectively. wavelengths),increasedpenetration,andshortmeas- The chemical structure of Cy is shown supplement- uringtimeduetotheappreciablephotonflux,hence ary figure S1(b). The measurements were performed providinganelegantsolutionformaterialscharacter- directly in solution. In the case of EC graphene, the isation.Inparticular,thex-raypairdistributionfunc- expandedgraphitepowder(ECpowder)wasalsocol- tion (XPDF) [24] can provide quantitative informa- lected and measured in solid form. Commercially tiononthecrystalstructure,e.g.theaveragedistances available graphite was also measured as a reference. between the neighbouring atoms, enabling insights The graphene concentration is determined by UV– into the material structure with nanoscale resolu- vis spectroscopy using an absorption coefficient of −1 −1 −1 −1 tion. Nevertheless, to the best of our knowledge, 2207lg m [32]and2460lg m [11]forEC these types of measurements have been rarely per- andLPEgraphene,respectively,measuredat660nm. formed on graphene-based materials. Previous stud- The concentrations are reported in table S1, in the ies with high energy x-rays [25–28] focussed on the supplementaryinformation. studyofvarioustypesofgraphiticcarbons,including The graphene nanosheets have been character- graphene oxide, specifically with the aim to identify ised by atomic force microscopy (AFM), Raman thetypeofdefects. spectroscopy, high resolution transmission electron Herein,weappliedtheXPDFtothecharacterisa- microscopy (HRTEM) and x-ray photoelectron tionofsolution-processedgrapheneproducedbyLPE spectroscopy (XPS). The thickness and lateral size and EC and dispersed in different solvents. The geo- distributions of both LPE and EC graphene flakes metricalarrangementofcarbonatomsintheringhas were estimated by AFM (figures 2(A) and (D)). been obtained from the analysis of the XPDF data. Figures 2(B) and (E) show the statistics of the peak Weobservestructuraldeformationsofthehexagonal thickness of the LPE and EC graphene, respectively, carbon ring, compared to the perfect planar geo- as extracted from AFM measurements, done over metry, inducing possible rippling at the scale of the 300 individual flakes, revealing that the dispersions inter-atomic distances, which could be related to the are mostly composed of thin (<10 layers) graphene interaction with the solvent molecules, as predicted flakes. The lateral size distribution of LPE and EC theoretically. grapheneinfigures2(C)and(F),respectively,shows 2 2DMater.10(2023)015006 ZYanetal Figure2.(A)TypicalAFMimageofLPEgraphenedrop-castedonSiO /Sisubstrate;(B)thicknessand(C)lateralsizedistribution ofLPEgraphenemeasuredbyAFM;(D)typicalAFMimageofECgraphene;(E)numberoflayersand(F)lateralsizedistribution ofECgraphenemeasuredbyAFM;(G)and(H)representativeRamanspectraofLPEandEGgraphenemeasuredwith514.5nm laser,respectively;(I)HRTEMimagesofLPEgraphene,insetshowsthecorrespondingFFTimage;(J)TEMimageoftheEC graphene,insetistheSAEDpattern;(K)and(L)XPSC1sspectraofLPEandECgraphene,respectively. that the as-prepared graphene nanosheets follows a LPE and EC, respectively. The typical Raman spec- broad distribution in size, where the average lateral trum of LPE and EC Gr shows the D and G peaks −1 −1 sizefortheLPEGrwas ∼223nmandbetween1and at ∼1350 cm and ∼1580 cm , respectively [33]. 4 µm for the EC Gr. The smaller flake size for the The D peak is activated by defects, but the specific LPEGrcomparedtotheECGrisduetotheextensive activation mechanism is different between the two sonication during the exfoliation process and is in samples. In the case of graphene produced by LPE, good agreement with our previous reports [7, 12]. theDpeakisactivatedbytheedgesofthenanosheets, Figures 2(G) and (H) shows representative Raman having lateral size comparable or smaller than that spectra measured on individual flakes produced by of the laser spot size [34]. The D peak in the Raman 3 2DMater.10(2023)015006 ZYanetal spectrumofECgraphene(figure2(H))islikelyactiv- structurefunction,S(Q),representingthenormalised ated by introducing functional groups during the scattering cross-section form [21]. A highly mono- electrochemical treatment, as evidenced by the XPS chromatic synchrotron x-ray beam was used; hence results. It should also be noted that the gas bubbles zero bremsstrahlung contribution can be assumed. −1 collapseontheelectrodescouldformdifferentkinds The top-hat width for the Lorch function of 1.0 Å ofdefectsinthegraphenebasalplane,includinggen- was utilised, and a minimum Fourier filter radius of −1 erating some vacancies. These defects could also be 1.25Å wasusedfortheFouriertransform. a result of structural defects introduced by the gas Figure3(B)showstheS(Q)ofallthesamplesafter evolution between the layers caused by the repeated subtractionofthebackground.Inthecaseofgraphite, ion intercalation/deintercalation process during the the intense peak-like features are characteristic of a exfoliationprocess. three-dimensionalandhighlycrystallinestructure.In −1 High resolution TEM (HRTEM) images reveal the case of EC powder, the peaks over Q = 3 Å that the graphene nanosheets produced by the LPE decrease in intensity, while the intensity of the peak −1 and EC processes are clean and of high crystallin- just below Q = 2 Å increases. The solution- ity. Moreover, both the FFT and SAED patterns of processed graphene obtained by LPE and by cath- LPE and EC graphene (inset of figures 2(I) and odic EC show a similar spectrum, characterised by a −1 (J), respectively) reveal the bright inner ring of broad peak at about Q = 1.1 Å . The missing/de- {0–110} spots and faint outer ring of {1–210} spots, creasedintensityofthepeaksisrelatedtoachangein in agreement with the typical diffraction pattern structure, associated with a reduced ordered config- of monolayer graphene. Moreover, the High-Angle urationoftheatoms.TheprocessedXPDFspectraup Annular Dark Field (HAADF) and the correspond- to25Åisshowninfigure3(C).Graphiteshowsobvi- ing energy disperse spectroscopy (EDS) mapping of ous atomic correlations up to 15 Å or more. In the thegrapheneflake(supplementaryfigureS2)demon- case of the graphite powder, the correlation is visible strate that graphene produced by the LPE process upto10Å.Ontheotherhand,inthecaseofsolution- consistsofpristineflakeswithoutanyimpurities. processed graphene, long-distance correlations are The chemical composition of the as-prepared completely missing. The absence of long-distance graphene was further investigated by the XPS. As correlation peaks indicates the disappearance of the showninfigures2(J)and(K),boththeXPSspectraof three-dimensional (3D) structure due to the exfoli- theLPEandECgrapheneshowasymmetricC1speak ation (as in the case of LPE and EC samples). The centred ∼284 eV corresponding to sp C–C bond. exfoliation caused lack of repeated atomic correla- Noticeably, only a tiny amount of oxygen related tion between the layers of graphite (i.e. loss of AB- functional groups (i.e. C–OH and C=O groups at stacking),asinthecaseoftheECpowder. 285.5 eV and 288.4 eV, respectively) were observed In the case of solution-processed graphene, only in LPE Gr, which mainly inherited form the start- the two peaks at short correlation distances (<5 Å) inggraphiteusedfortheexfoliationprocess[35].On are clearly visible. However, a closer look in this theotherhand,theoxygencontentoftheECGrwas region, figure 4(A), shows that the number of peaks foundtobearound7.8at%comparedtothe5.5at% andtheirpositionsisslightlydifferent,dependingon of pure graphite (supplementary figures S3 and S4), the sample considered. These peaks are very import- supporting the assumption the EC exfoliation pro- ant because their position is associated with the C– cessusedislargelynonoxidative.Therefore,boththe C atomic distance in the ring: by using the carbon LPE and EC Gr used in this study are of high quality hexagon model from graphite in figure 4(B), one and crystallinity as supported by different character- wouldexpecttoseefivepeaksatthedistances1.41Å, izationmethodsdiscussedabove. 2.42 Å, 2.86 Å, 3.74 Å, 4.22 Å, corresponding to the For x-ray characterization, the samples were C1–C2, C1–C3, CI–C4, C1–C5, C1–C6 atomic dis- packed in a borosilicate capillary with 1.5 mm tances, respectively. This is well observed with the diameter. The background signal measurement is selected Q range of XPDF spectrum, as shown in conducted, (as shown in supplementary figure S5) figure4(A),whichalsoagreeswithpreviousresearch and subtracted from the signal of the graphene. As [36,37]. shown in figure 3(A), during the synchrotron x-ray Figure 4(C) shows that the C1–C2 bonding in experiments, the capillaries are mounted horizont- graphite powder increases up to 1.46 Å, while it ally and rotate about its long axis centring the x- is fixed at 1.43 Å or 1.44 Å for solution-processed ray beam at the upper half of the capillary to avoid graphene,irrespectiveoftheexfoliationmethodused. any precipitations of the highly concentrated solu- Notethatcathodicgraphenehasasizecomparableto tion sample. The exposure time for each x-ray scat- graphene flakes made by LPE [12]. The C1–C3 dis- tering measurement was 300 s. The experimentally tanceslightlyincreasesingraphitepowdercompared collecteddiffractionintensitydata,includingBragg’s to bulk graphite, while it is much smaller in the case scattering signal and high Q range scattering signal ofsolution-processedgraphene.Thesameisobserved can be primarily processed into the total scattering forC1–C4,althoughthesignalofsolution-processed 4 2DMater.10(2023)015006 ZYanetal Figure3.(A)Schematicillustrationofthesynchrotronx-rayexperimentsetupwith2Dareadetector;(B)structurefactorresults calculatedfromthecollecteddiffractionintensity;(C)XPDFresultsofLPE,ECexfoliatedgrapheneandBulkgraphitesamplesup to25Å.TheXPDFresultisobtainedbyFouriertransformationofstructurefactor. grapheneisratherweakwithabroadpeakbutcanstill means no constructive scattering can occur even berecognisedatabout2.85Å,whichindicatesthatthe without removing the background. Therefore, all C1–C4 atomic correlation is hardly visible measured of the characteristic peaks of the solvent molecules as a fixed value. In the case of the C1–C5 distance, are not observed in our spectra. All the correlation no significant differences are observed between bulk peaks, and in particular the one at 1.43–1.44 Å for graphite and EC graphite powder. In contrast, the solution-processed graphene, are doubtlessly com- C1–C5 distances are reduced in the case of solution- ing from graphene. Thus, the difference between processedgraphene,withnodependenceontheexfo- graphene in NMP and Cyrene is minimal and can liation method. The peak associated with C1–C6 is beneglected[29]. missing in the case of solution-processed graphene. The XPDF results confirm that cathodic EC and The correlation of longer distance C1–C6 becomes LPE can provide nanosheets with thickness small too weak to be detected by the XPDF, which can be enough to be considered two-dimensional. How- regardedasuncorrelatedpair. ever,slightlydifferentgeometricarrangementsofthe Interestingly, any correlation associated with atoms in the hexagonal ring have been observed as NMP or Cyrene (for structure details readers are compared to graphite. Figure 4(C) shows the result- referred to [29]) is not observed in solution- ing carbon hexagon models for solution-processed processed graphene. The intermolecular correla- graphene in NMP and EC graphite powder com- tion of the solvent molecules could be visible if pared to that of graphite. Our results show that ion the molecules are ordered in space. However, the intercalationandrelatedlayers’expansion,causedby NMP and Cyrene molecules are probably randomly thecathodicECprocess,givesrisetoasmallringdis- spaced with relatively low concentration, which tortioncomparedtopristinegraphene.However,the 5 2DMater.10(2023)015006 ZYanetal Figure4.(A)XPDFresultsofacarbonhexagonrangeupto4.5Å.(B)Carbonhexagonmodelfromgraphite.Distancesbetween atomsarenoted.(C)Carbonhexagonmodelswithdistortionofsinglelayergrapheneinbulkgraphite,ECexpandedgraphite powder,graphenedispersioninNMPsolution. ◦ ◦ ring distortion further increases when the material calculated as 111.9 and 114 for graphene disper- is dispersed in a solvent. Remarkably, as both LPE sion and powder. The normal surface angle to the and EC graphene dispersed in the same solvent sheet of the powder sample should be less than 6 ◦ ◦ show exactly the same type of distortion of the (i.e.theangledeviationof114 from120 ),inagree- hexagonalring,thiseffectislikelytobecausedbythe ment with the angle deviation from Meyer’s work interactionbetweensolventmoleculesandgraphene. [38],wheretheangledeviationfromthesheetis±5 . Nevertheless, further research needs to done to fully Likely, this effect is caused by the interaction of the confirmthisobservation. solventwithgraphene. Meyeretal[38]havesuggestedripplesinsuspen- ded graphene through TEM investigations, as con- firmed by other groups [39–41]. In particular, an 3.Conclusions average of 0.7 Å height fluctuation (i.e. ripple height normaltothesheet)wasfoundbyusingMonteCarlo Our work presents the first characterisation of simulation [42]. The bond length deviation is pre- solution-processed graphene in NMP and Cyrene dicted from 1.31 Å to 1.54 Å. This could include the by high energy x-ray scattering and related XPDF short double bond of 1.31 Å, a conjugated bond of analysis. The results show the disappearance of 1.42Å,anduptoalongsinglebondof1.54Å.Hence, long-range atomic correlations, confirming the pro- the three bonds of each carbon atom within the duction of 2D nanosheets and that the hexagonal hexagoncouldbedifferent.Ourresultsshowringdis- atomicstructureisstronglydistortedwhengraphene tortionscompatiblewiththeproposedmodelforsus- is suspended in a medium. In particular, the first pendedgraphene[39];astheC1–C3atomicdistance C–C distance slightly increases, while the second ofsolution-processedgraphenebecomesshorterthan C–Cdistancedecreases,resultinginadistortionthat the bulk graphite. However, it needs to be noted couldbecompatiblewithripplesformationobserved that TEM measurements were done on the suspen- in suspended graphene, likely to be caused by the dedgraphenewhileourmeasurementswereobtained interaction of solvent molecules with the graphene forgrapheneinsolution.Theincreasingbondlengths nanosheets.Ourresultsdemonstratethepotentialof differences indicate that the graphene’s increased XPDFasapowerfultooltocharacterise2Dmaterials density is not perfectly planar when dispersed in a andacquirequantitativestructuralinformationwith solvent. The bonding angle between C1–C2–C3 is atomicresolution. 6 2DMater.10(2023)015006 ZYanetal 4.Methods were used for the measurements. The x-ray pho- toelectron spectroscopy (XPS) measurements were 4.1.Materials performed using the K-Alpha x-ray Photoelectron High purity graphite foil (99.8% metal basis), Spectrometer (XPS) System from Thermo Scientific. graphite rod (99.99% metal basis) and ammonium The photon source was a monochromatized Al Kα sulfate ((NH ) SO , 98+%) were purchased line (hν = 1486.6 eV). The spectra were acquired 4 2 4 from Alfa Aesar. Anhydrous dimethyl sulfoxide usingaspotsizeof300 µmandconstantpassenergy (DMSO) (99.9%), graphite flakes (100+ mesh), (150 eV for survey and 20 eV for high resolution 1-Pyrenesulfonic acid sodium salt (PS1) and iso- spectra). A flood gun with combined electrons and propyl alcohol (IPA) were purchased from Sigma- lowenergyArionsisusedduringthemeasurements. Aldrich. Isomolded graphite (>99.95%) rods were HRTEM images were acquired on a JEOL 2100-F purchased from GraphiteStore. Cesium perchlorate microscope with a field-emission gun operated at (99%) was obtained from Fisher Scientific. The nat- 200 kV accelerating voltage providing direct images ural kish graphite was bought from Graphexel Ltd. of the atomic structure. A HAADF detector and an Allthechemicalsandmaterialswereusedasreceived. Oxford high solid-angle silicon drift detector x-ray energy dispersive spectrometer (EDS) system was 4.2.Exfoliation usedforchemicalelementalanalysis. The LPE graphene in NMP was prepared by adding 4.4.X-rayscattering 300 mg of graphite flakes into 100 ml of NMP, fol- lowed by sonicating the mixture at 600 W using The synchrotron x-ray scattering experiments were conductedinI15–1beamline,DiamondLightSource, Hilsonic bath sonicator for 5 d. Afterwards, the dis- persion was centrifuged using a Sigma 1–14k refri- U.K. The monochromatic x-ray beam with 76.7 keV (wavelength of 0.161669 Å) was employed [43]. A gerated centrifuge at 903 g for 20 min to remove un-exfoliatedgraphite.Toobtainhighlyconcentrated 2D Perkin Elmer XRD detector with active area of 2, 409.6 × 409.6 mm and pixel size of 100 µm was graphene, the dispersion was further centrifuged at 16600 g for 1 h., followed by re-dispersing the sed- applied close tosample to providelarge Q rangeand high-quality scattering data. Here Q = (4π sinθ) ⁄λ, imented graphene into a small volume of NMP. The ECispreparedbythecathodicelectrochemicalexfoli- where λ is the wavelength and 2θ is the angle between incident and scattered x-rays. The collec- ationasdescribedinourpreviouswork[12].Briefly, a pellet of natural graphite is used as a cathode, and teddiffractionintensitydataisprocessedbysoftware GudrunX [44] which subtracts the self-scattering Pt mesh was used as the anode. The electrolyte was 1M lithium chloride (Sigma Aldrich, 99.9%), and intensity, Compton scattering and multiple scatter- ing, etc. Then, the total scattering structure factor, Triethylamine hydrochloride in dimethyl sulfoxide. The exfoliation products were washed to remove the S(Q)andXPDF,G(r)areobtainedas; electrolyte with water and ethanol until the pH was I(Q) neutral,andtheproductswereseparatedbyfiltration S (Q) = (1) usingAnodiscaluminamembraneswith100nmpore size and then dried at 200 C under Ar atmosphere. max Thedrypowderwasthendispersedinasmallamount G(r) = Q[S(Q)−1]sin(Qr)dQ (2) ofNMA,asintheLPEsamples. π min 4.3.Materialscharacterization where I(Q) is the collected and processed diffraction UV–Vis spectroscopy of the graphene dispersions intensity. The coherent single-scattering intensity is were measured by using a PerkinElmer I-900 UV– desired; b is the element scattering amplitude (f is Vis–NIR spectrometer. A Bruker Atomic Froce usedforx-rayscattering);<…>denotesanaveraging Microscope (MultiMode 8) in Peak Force Tapping process.Fordetailedtheory,[21,45]isreferred. mode, equipped with ScanAsyst-Air tips is used to determine the lateral size and thickness distribution Dataavailabilitystatement of the graphene flakes. The samples were prepared by drop casting the dispersion on a clean silicon All data that support the findings of this study are substrate; several hundreds of individual flakes were included within the article (and any supplementary selected, after complete solvent evaporation, for lat- files). eral size and thickness analysis. The same sample preparationhasbeenusedforRamanmeasurements. Acknowledgments Raman measurements were performed using a Ren- ishaw Invia Raman spectrometer equipped with a We acknowledged Diamond Light Source for grant- 514.5 nm excitation line with 1 mW laser power. ing beamtime at I15-1 (CY24816). W M acknow- 100× NA0.85 objective lens, giving a spatial resol- ledges the funding from EPSRC (UK) Grant −1 ution of ∼500 nm, and 2400 grooves nm grating EP/P02680X/1. The work of M J G G is supported 7 2DMater.10(2023)015006 ZYanetal bytheXuntadeGalicia(Spain)PostdoctoralFellow- [8] TorrisiFetal2012Inkjet-printedgrapheneelectronicsACS Nano62992–3006 ship with reference ED481B-2019-015. C C and K P [9] NicolosiV,ChhowallaM,KanatzidisMG,StranoMSand acknowledgetheGrapheneFlagshipCore3(Contract ColemanJN2013Liquidexfoliationoflayeredmaterials No. 881603) and the ERC Project PEP2D (Contract Science3401226419 No. 770047). O R acknowledges financial support [10] BackesC,HigginsTM,KellyA,BolandC,HarveyA, HanlonDandColemanJN2017Guidelinesforexfoliation, from the Lloyd’s Register Foundation. C D and J U characterizationandprocessingoflayeredmaterials acknowledge the Office of Science, Office of Basic producedbyliquidexfoliationChem.Mater.29243–55 Energy Sciences, of the U.S. Department of Energy [11] HernandezYetal2008High-yieldproductionofgraphene (ContractNo.DE-AC02-05CH11231). byliquid-phaseexfoliationofgraphiteNat.Nanotechnol. 3563–8 [12] AbdelkaderAM,KinlochIAandDryfeRAW2014 Authorcontributions Continuouselectrochemicalexfoliationofmicrometer-sized grapheneusingsynergisticionintercalationsandorganic solventsACSAppl.Mater.Interfaces61632–9 Z Y, A A and W M conceived the approach and [13] AbdelkaderAM2015Electrochemicalsynthesisofhighly designed the experiments. A A, W M, and TF car- corrugatedgraphenesheetsforhighperformance ried out the x-ray experiments while M G and K P supercapacitorsJ.Mater.Chem.A38519–25 prepared materials and carried out lab experiments. [14] ParvezK,WuZ-S,LiR,LiuX,GrafR,FengXandMüllenK 2014Exfoliationofgraphiteintographeneinaqueous O R, C D and J U performed the material character- solutionsofinorganicsaltsJ.Am.Chem.Soc.1366083–91 izations with AFM, HRTEM, EDS and XPS analysis. [15] CooperAJ,WilsonNR,KinlochIAandDryfeRAW2014 Z Y performed the analysis the data and drafted the Singlestageelectrochemicalexfoliationmethodforthe manuscriptwithinputsfromCC,AAandWM.All productionoffew-layergrapheneviaintercalationof tetraalkylammoniumcationsCarbon66340–50 authors discussed the results and contributed to the [16] YangY,LuF,ZhouZ,SongW,ChenQandJiX2013 manuscript. Electrochemicallycathodicexfoliationofgraphenesheetsin roomtemperatureionicliquidsN-butyl, ORCIDiDs methylpyrrolidiniumbis(trifluoromethylsulfonyl)imideand theirelectrochemicalpropertiesElectrochim.Acta1139–16 [17] WickPetal2014Classificationframeworkfor ZhengyuYan https://orcid.org/0000-0002-2543- graphene-basedmaterialsAngew.Chem.Int.Ed.537714–8 [18] KostarelosKandNovoselovKS2014Graphenedevicesfor MaríaJGGuimarey https://orcid.org/0000-0003- lifeNat.Nanotechnol.9744–5 [19] PollardAJetal2017Characterisationofthestructureof 0483-4136 graphene(availableat:https://eprintspublications.npl.co. KhaledParvez https://orcid.org/0000-0003-2851- uk/8654/) [20] ISO/TS21356–12021Nanotechnologies—Structural ChaochaoDun https://orcid.org/0000-0002- CharacterizationofGraphene—Part1:Graphenefrom PowdersandDispersions(InternationalOrganisationfor 3215-6478 Standardization)(availableatwww.iso.org/standard/70757. OliverRead https://orcid.org/0000-0001-9771- html) [21] EgamiTandBillingeSJLEds2012UnderneaththeBragg JeffreyJUrban https://orcid.org/0000-0003-4909- PeaksStructuralAnalysisofComplexMaterialsvol 16(Oxford:Pergamon) [22] BilderbackDH,ElleaumePandWeckertE2005Reviewof CinziaCasiraghi https://orcid.org/0000-0001- thirdandnextgenerationsynchrotronlightsourcesJ.Phys. 7185-0377 B:At.Mol.Opt.Phys.38S773–97 WajiraMirihanage https://orcid.org/0000-0002- [23] HastingsJB,ThomlinsonWandCoxDE1984Synchrotron x-raypowderdiffractionJ.Appl.Crystallogr.1785–95 9083-269X [24] WestneatMW,SochaJJandLeeW-K2008Advancesin biologicalstructure,function,andphysiologyusing References synchrotronx-rayimagingAnnu.Rev.Physiol.70119–42 [25] SmithMA,FoleyHCandLoboRF2004Asimplemodel [1] BonaccorsoF,BartolottaA,ColemanJNandBackesC2016 describesthePDFofanon-graphitisingcarbonCarbon 2D-crystal-basedfunctionalinksAdv.Mater.286136–66 422041–8 [2] BackesCetal2020Productionandprocessingofgraphene [26] PetkovV,DifrancescoRG,BillingeSJL,AcharyaMand andrelatedmaterials2DMater.7022001 FoleyHC1999LocalstructureofnanoporouscarbonsPhil. [3] CaiX,LuoY,LiuBandChengH-M2018Preparationof2D Mag.B791519–30 materialdispersionsandtheirapplicationsChem.Soc.Rev. [27] WoznicaN,HawelekL,FischerHE,BobrinetskiyIand 476224–66 BurianA2015Theatomicscalestructureofgraphene [4] CiesielskiAandSamor`ıP2014Grapheneviasonication powderstudiedbyneutronandx-raydiffractionJ.Appl. assistedliquid-phaseexfoliationChem.Soc.Rev.43381–98 Crystallogr.481429–36 [5] KellyAGetal2017All-printedthin-filmtransistorsfrom [28] Wo´znicaN,HawelekL,DuberS,FischerHE,HonkimäkiV, networksofliquid-exfoliatednanosheetsScience35669–73 PawlytaM,BulouAandBurianA2017Theatomicscale [6] SecorEBandHersamMC2015Emergingcarbonand structureofsaccharose-basedcarbonsPhil.Mag. post-carbonnanomaterialinksforprintedelectronicsJ. 971675–97 Phys.Chem.Lett.6620–6 [29] SalavagioneHJ,SherwoodJ,DeBruynM,BudarinVL, [7] McManusDetal2017Water-basedandbiocompatible2D EllisGJ,ClarkJHandShuttleworthPS2017Identification crystalinksforall-inkjet-printedheterostructuresNat. ofhighperformancesolventsforthesustainableprocessing Nanotechnol.12343–50 ofgrapheneGreenChem.192550–60 8 2DMater.10(2023)015006 ZYanetal [30] ColemanJN2013Liquidexfoliationofdefect-freegraphene [38] MeyerJC,GeimAK,KatsnelsonMI,NovoselovKS, Acc.Chem.Res.4614–22 BoothTJandRothS2007Thestructureofsuspended [31] WangJ,MangaKK,BaoQandLohKP2011High-yield graphenesheetsNature44660–63 synthesisoffew-layergrapheneflakesthrough [39] AmiriA,NaraghiM,AhmadiG,SoleymanihaMand electrochemicalexpansionofgraphiteinpropylene ShanbediM2018Areviewonliquid-phaseexfoliationfor carbonateelectrolyteJ.Am.Chem.Soc.1338888–91 scalableproductionofpuregraphene,wrinkled,crumpled [32] ParvezK,WorsleyR,AlievaA,FeltenAandCasiraghiC andfunctionalizedgrapheneandchallengesFlatChem 2019Water-basedandinkjetprintableinksmadeby 840–71 electrochemicallyexfoliatedgrapheneCarbon149213–21 [40] LiZ,KinlochIA,YoungRJ,NovoselovKS, [33] NagyteVetal2020Ramanfingerprintsofgraphene AnagnostopoulosG,PartheniosJ,GaliotisC,PapagelisK, producedbyanodicelectrochemicalexfoliationNanoLett. LuC-YandBritnellL2015Deformationofwrinkled 203411–9 grapheneACSNano93917–25 [34] CasiraghiC,HartschuhA,QianH,PiscanecS,GeorgiC, [41] LiuN,PanZ,FuL,ZhangC,DaiBandLiuZ2011The FasoliA,NovoselovKS,BaskoDMandFerrariAC2009 originofwrinklesontransferredgrapheneNanoRes. RamanspectroscopyofgrapheneedgesNanoLett.91433–41 4996–1004 [35] ShangNG,PapakonstantinouP,SharmaS,LubarskyG, [42] FasolinoA,LosJHandKatsnelsonMI2007Intrinsic LiM,McNeillDW,QuinnAJ,ZhouWandBlackleyR2012 ripplesingrapheneNat.Mater.6858–61 Controllableselectiveexfoliationofhigh-qualitygraphene [43] ChaterPA,KeebleD,WharmbyM,SpainT,FilikJand nanosheetsandnanodotsbyionicliquidassistedgrinding WilhelmH2017TheautomatedXPDFbeamlineatdiamond Chem.Commun.481877 lightsourceActaCrystallogr.A73C69 [36] PetkovV,RenY,KabekkoduSandMurphyD2013Atomic [44] KeenDA2001Acomparisonofvariouscommonlyused pairdistributionfunctionsanalysisofdisorderedlow-Z correlationfunctionsfordescribingtotalscatteringJ.Appl. materialsPhys.Chem.Chem.Phys.158544 Crystallogr.34172–7 [37] BurianA,DoreJCandJurkiewiczK2019Structuralstudies [45] BillingeSJL2019Theriseofthex-rayatomicpair ofcarbonsbyneutronandx-rayscatteringRep.Prog.Phys. distributionfunctionmethod:aseriesoffortunateevents 82016501 Phil.Trans.R.Soc.A37720180413 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png 2D Materials IOP Publishing

Nanoscale insights into the structure of solution-processed graphene by x-ray scattering

Loading next page...
 
/lp/iop-publishing/nanoscale-insights-into-the-structure-of-solution-processed-graphene-nwnsWX9UTM

References (47)

Publisher
IOP Publishing
Copyright
© 2022 The Author(s). Published by IOP Publishing Ltd
eISSN
2053-1583
DOI
10.1088/2053-1583/ac9b6f
Publisher site
See Article on Publisher Site

Abstract

Chemicalexfoliationisanattractiveapproachforthesynthesisofgrapheneduetoitslowcostand simplicity.However,challengesstillremaininthecharacterizationofsolution-processedgraphene, inparticularwithatomicresolution.Throughthisworkwedemonstratethex-raypairdistribution functionasanovelapproachtostudysolution-processedgrapheneorother2Dmaterialswith atomicresolution,directlyinsolution,producedbyliquid-phaseandelectrochemicalexfoliations. Theresultsshowthedisappearanceoflong-rangeatomiccorrelations,inbothcases,confirming theproductionofsingleandfew-layergraphene.Inaddition,aconsiderableringdistortionhas beenobservedascomparedtographite,irrespectiveofthesolventused:thenormalsurfaceangle tothesheetofthepowdersampleshouldbelessthan6 ,compatiblewithripplesformation observedinsuspendedgraphene.Weattributethiseffecttotheinteractionofsolventmolecules withthegraphenenanosheets. 1.Introduction the intercalated ions, the graphite electrode works as an anode or cathode, hosting oxidation or reduction One of the most attractive and industrially scal- reactions,respectively[12–14].CathodicECgivesrise able methods for graphene production is given to defect-free graphene, but exfoliation does require by chemical exfoliation of graphite [1–4]. This several hours [15, 16], while anodic EC is quick but approach gives rise to graphene dispersions, which givesrisetoslightlyoxidisedgraphene[14]. can be further processed with simple and low-cost Despite the use of simple methods to pro- methods such as drop casting and inkjet printing duce solution-processed graphene, its character- [5–8]. The most used chemical exfoliation isation is very challenging because the nanosheets approaches are liquid phase exfoliation (LPE) and come in different sizes, thicknesses and chemical electrochemical exfoliation (EC). The LPE method functionalisation [10]. In addition, the lack of met- relies on the use of ultrasound and/or shear force to rology standards strongly limits the commercial- exfoliate bulk graphite into graphene suspended in isation of graphene-based products [10, 17, 18]. a suitable solvent, such as N-Methyl-2-pyrrolidone Only recently, the community has defined some (NMP), N,N-dimethylformamide (DMF), and guidelinesforthecharacterisationofgraphene-based Cyrene (Cy) [9–11]. The EC process is based on materials [19], which led to the development of the expanding the graphite layers following the inter- first ISO/IEC standard (ISO/TS 21356-1:2021) for calation of ions and small molecules driven by an measuring the structural properties of graphene external electric field. According to the charge of [20]. In particular, information on the atomic ©2022TheAuthor(s). PublishedbyIOPPublishingLtd 2DMater.10(2023)015006 ZYanetal Figure1.Schematicillustrationoftheexfoliationprocess:(A)sonication-aidedLPEprocessfollowedbycentrifugation;(B)EC processusinggraphiteascathodicelectrode. structureofsolution-processedgrapheneiscurrently 2.Resultsanddiscussion providedbytransmissionelectronmicroscopy,which requires specific sample preparation, and it is time- Solution-processed graphene was produced by using consuming,soitcanonlybeperformedonaselected cathodic EC and LPE methods, as outlined in number of nanosheets. The electron beam can also figures 1(A) and (B), (and supplementary figure damageorchangethestructureofthenanosheets. S1(a)) following the approaches reported previously Synchrotron x-ray based characterisation tech- [12, 29–31]. The solution process graphene in either niques are increasingly employed in material science NMP and Cy are named as Gr LPE (NMP) Gr LPE [21–24]becauseoftheirhighphotonenergy(shorter (Cy), Gr EC (NMP) and Gr EC (Cy), respectively. wavelengths),increasedpenetration,andshortmeas- The chemical structure of Cy is shown supplement- uringtimeduetotheappreciablephotonflux,hence ary figure S1(b). The measurements were performed providinganelegantsolutionformaterialscharacter- directly in solution. In the case of EC graphene, the isation.Inparticular,thex-raypairdistributionfunc- expandedgraphitepowder(ECpowder)wasalsocol- tion (XPDF) [24] can provide quantitative informa- lected and measured in solid form. Commercially tiononthecrystalstructure,e.g.theaveragedistances available graphite was also measured as a reference. between the neighbouring atoms, enabling insights The graphene concentration is determined by UV– into the material structure with nanoscale resolu- vis spectroscopy using an absorption coefficient of −1 −1 −1 −1 tion. Nevertheless, to the best of our knowledge, 2207lg m [32]and2460lg m [11]forEC these types of measurements have been rarely per- andLPEgraphene,respectively,measuredat660nm. formed on graphene-based materials. Previous stud- The concentrations are reported in table S1, in the ies with high energy x-rays [25–28] focussed on the supplementaryinformation. studyofvarioustypesofgraphiticcarbons,including The graphene nanosheets have been character- graphene oxide, specifically with the aim to identify ised by atomic force microscopy (AFM), Raman thetypeofdefects. spectroscopy, high resolution transmission electron Herein,weappliedtheXPDFtothecharacterisa- microscopy (HRTEM) and x-ray photoelectron tionofsolution-processedgrapheneproducedbyLPE spectroscopy (XPS). The thickness and lateral size and EC and dispersed in different solvents. The geo- distributions of both LPE and EC graphene flakes metricalarrangementofcarbonatomsintheringhas were estimated by AFM (figures 2(A) and (D)). been obtained from the analysis of the XPDF data. Figures 2(B) and (E) show the statistics of the peak Weobservestructuraldeformationsofthehexagonal thickness of the LPE and EC graphene, respectively, carbon ring, compared to the perfect planar geo- as extracted from AFM measurements, done over metry, inducing possible rippling at the scale of the 300 individual flakes, revealing that the dispersions inter-atomic distances, which could be related to the are mostly composed of thin (<10 layers) graphene interaction with the solvent molecules, as predicted flakes. The lateral size distribution of LPE and EC theoretically. grapheneinfigures2(C)and(F),respectively,shows 2 2DMater.10(2023)015006 ZYanetal Figure2.(A)TypicalAFMimageofLPEgraphenedrop-castedonSiO /Sisubstrate;(B)thicknessand(C)lateralsizedistribution ofLPEgraphenemeasuredbyAFM;(D)typicalAFMimageofECgraphene;(E)numberoflayersand(F)lateralsizedistribution ofECgraphenemeasuredbyAFM;(G)and(H)representativeRamanspectraofLPEandEGgraphenemeasuredwith514.5nm laser,respectively;(I)HRTEMimagesofLPEgraphene,insetshowsthecorrespondingFFTimage;(J)TEMimageoftheEC graphene,insetistheSAEDpattern;(K)and(L)XPSC1sspectraofLPEandECgraphene,respectively. that the as-prepared graphene nanosheets follows a LPE and EC, respectively. The typical Raman spec- broad distribution in size, where the average lateral trum of LPE and EC Gr shows the D and G peaks −1 −1 sizefortheLPEGrwas ∼223nmandbetween1and at ∼1350 cm and ∼1580 cm , respectively [33]. 4 µm for the EC Gr. The smaller flake size for the The D peak is activated by defects, but the specific LPEGrcomparedtotheECGrisduetotheextensive activation mechanism is different between the two sonication during the exfoliation process and is in samples. In the case of graphene produced by LPE, good agreement with our previous reports [7, 12]. theDpeakisactivatedbytheedgesofthenanosheets, Figures 2(G) and (H) shows representative Raman having lateral size comparable or smaller than that spectra measured on individual flakes produced by of the laser spot size [34]. The D peak in the Raman 3 2DMater.10(2023)015006 ZYanetal spectrumofECgraphene(figure2(H))islikelyactiv- structurefunction,S(Q),representingthenormalised ated by introducing functional groups during the scattering cross-section form [21]. A highly mono- electrochemical treatment, as evidenced by the XPS chromatic synchrotron x-ray beam was used; hence results. It should also be noted that the gas bubbles zero bremsstrahlung contribution can be assumed. −1 collapseontheelectrodescouldformdifferentkinds The top-hat width for the Lorch function of 1.0 Å ofdefectsinthegraphenebasalplane,includinggen- was utilised, and a minimum Fourier filter radius of −1 erating some vacancies. These defects could also be 1.25Å wasusedfortheFouriertransform. a result of structural defects introduced by the gas Figure3(B)showstheS(Q)ofallthesamplesafter evolution between the layers caused by the repeated subtractionofthebackground.Inthecaseofgraphite, ion intercalation/deintercalation process during the the intense peak-like features are characteristic of a exfoliationprocess. three-dimensionalandhighlycrystallinestructure.In −1 High resolution TEM (HRTEM) images reveal the case of EC powder, the peaks over Q = 3 Å that the graphene nanosheets produced by the LPE decrease in intensity, while the intensity of the peak −1 and EC processes are clean and of high crystallin- just below Q = 2 Å increases. The solution- ity. Moreover, both the FFT and SAED patterns of processed graphene obtained by LPE and by cath- LPE and EC graphene (inset of figures 2(I) and odic EC show a similar spectrum, characterised by a −1 (J), respectively) reveal the bright inner ring of broad peak at about Q = 1.1 Å . The missing/de- {0–110} spots and faint outer ring of {1–210} spots, creasedintensityofthepeaksisrelatedtoachangein in agreement with the typical diffraction pattern structure, associated with a reduced ordered config- of monolayer graphene. Moreover, the High-Angle urationoftheatoms.TheprocessedXPDFspectraup Annular Dark Field (HAADF) and the correspond- to25Åisshowninfigure3(C).Graphiteshowsobvi- ing energy disperse spectroscopy (EDS) mapping of ous atomic correlations up to 15 Å or more. In the thegrapheneflake(supplementaryfigureS2)demon- case of the graphite powder, the correlation is visible strate that graphene produced by the LPE process upto10Å.Ontheotherhand,inthecaseofsolution- consistsofpristineflakeswithoutanyimpurities. processed graphene, long-distance correlations are The chemical composition of the as-prepared completely missing. The absence of long-distance graphene was further investigated by the XPS. As correlation peaks indicates the disappearance of the showninfigures2(J)and(K),boththeXPSspectraof three-dimensional (3D) structure due to the exfoli- theLPEandECgrapheneshowasymmetricC1speak ation (as in the case of LPE and EC samples). The centred ∼284 eV corresponding to sp C–C bond. exfoliation caused lack of repeated atomic correla- Noticeably, only a tiny amount of oxygen related tion between the layers of graphite (i.e. loss of AB- functional groups (i.e. C–OH and C=O groups at stacking),asinthecaseoftheECpowder. 285.5 eV and 288.4 eV, respectively) were observed In the case of solution-processed graphene, only in LPE Gr, which mainly inherited form the start- the two peaks at short correlation distances (<5 Å) inggraphiteusedfortheexfoliationprocess[35].On are clearly visible. However, a closer look in this theotherhand,theoxygencontentoftheECGrwas region, figure 4(A), shows that the number of peaks foundtobearound7.8at%comparedtothe5.5at% andtheirpositionsisslightlydifferent,dependingon of pure graphite (supplementary figures S3 and S4), the sample considered. These peaks are very import- supporting the assumption the EC exfoliation pro- ant because their position is associated with the C– cessusedislargelynonoxidative.Therefore,boththe C atomic distance in the ring: by using the carbon LPE and EC Gr used in this study are of high quality hexagon model from graphite in figure 4(B), one and crystallinity as supported by different character- wouldexpecttoseefivepeaksatthedistances1.41Å, izationmethodsdiscussedabove. 2.42 Å, 2.86 Å, 3.74 Å, 4.22 Å, corresponding to the For x-ray characterization, the samples were C1–C2, C1–C3, CI–C4, C1–C5, C1–C6 atomic dis- packed in a borosilicate capillary with 1.5 mm tances, respectively. This is well observed with the diameter. The background signal measurement is selected Q range of XPDF spectrum, as shown in conducted, (as shown in supplementary figure S5) figure4(A),whichalsoagreeswithpreviousresearch and subtracted from the signal of the graphene. As [36,37]. shown in figure 3(A), during the synchrotron x-ray Figure 4(C) shows that the C1–C2 bonding in experiments, the capillaries are mounted horizont- graphite powder increases up to 1.46 Å, while it ally and rotate about its long axis centring the x- is fixed at 1.43 Å or 1.44 Å for solution-processed ray beam at the upper half of the capillary to avoid graphene,irrespectiveoftheexfoliationmethodused. any precipitations of the highly concentrated solu- Notethatcathodicgraphenehasasizecomparableto tion sample. The exposure time for each x-ray scat- graphene flakes made by LPE [12]. The C1–C3 dis- tering measurement was 300 s. The experimentally tanceslightlyincreasesingraphitepowdercompared collecteddiffractionintensitydata,includingBragg’s to bulk graphite, while it is much smaller in the case scattering signal and high Q range scattering signal ofsolution-processedgraphene.Thesameisobserved can be primarily processed into the total scattering forC1–C4,althoughthesignalofsolution-processed 4 2DMater.10(2023)015006 ZYanetal Figure3.(A)Schematicillustrationofthesynchrotronx-rayexperimentsetupwith2Dareadetector;(B)structurefactorresults calculatedfromthecollecteddiffractionintensity;(C)XPDFresultsofLPE,ECexfoliatedgrapheneandBulkgraphitesamplesup to25Å.TheXPDFresultisobtainedbyFouriertransformationofstructurefactor. grapheneisratherweakwithabroadpeakbutcanstill means no constructive scattering can occur even berecognisedatabout2.85Å,whichindicatesthatthe without removing the background. Therefore, all C1–C4 atomic correlation is hardly visible measured of the characteristic peaks of the solvent molecules as a fixed value. In the case of the C1–C5 distance, are not observed in our spectra. All the correlation no significant differences are observed between bulk peaks, and in particular the one at 1.43–1.44 Å for graphite and EC graphite powder. In contrast, the solution-processed graphene, are doubtlessly com- C1–C5 distances are reduced in the case of solution- ing from graphene. Thus, the difference between processedgraphene,withnodependenceontheexfo- graphene in NMP and Cyrene is minimal and can liation method. The peak associated with C1–C6 is beneglected[29]. missing in the case of solution-processed graphene. The XPDF results confirm that cathodic EC and The correlation of longer distance C1–C6 becomes LPE can provide nanosheets with thickness small too weak to be detected by the XPDF, which can be enough to be considered two-dimensional. How- regardedasuncorrelatedpair. ever,slightlydifferentgeometricarrangementsofthe Interestingly, any correlation associated with atoms in the hexagonal ring have been observed as NMP or Cyrene (for structure details readers are compared to graphite. Figure 4(C) shows the result- referred to [29]) is not observed in solution- ing carbon hexagon models for solution-processed processed graphene. The intermolecular correla- graphene in NMP and EC graphite powder com- tion of the solvent molecules could be visible if pared to that of graphite. Our results show that ion the molecules are ordered in space. However, the intercalationandrelatedlayers’expansion,causedby NMP and Cyrene molecules are probably randomly thecathodicECprocess,givesrisetoasmallringdis- spaced with relatively low concentration, which tortioncomparedtopristinegraphene.However,the 5 2DMater.10(2023)015006 ZYanetal Figure4.(A)XPDFresultsofacarbonhexagonrangeupto4.5Å.(B)Carbonhexagonmodelfromgraphite.Distancesbetween atomsarenoted.(C)Carbonhexagonmodelswithdistortionofsinglelayergrapheneinbulkgraphite,ECexpandedgraphite powder,graphenedispersioninNMPsolution. ◦ ◦ ring distortion further increases when the material calculated as 111.9 and 114 for graphene disper- is dispersed in a solvent. Remarkably, as both LPE sion and powder. The normal surface angle to the and EC graphene dispersed in the same solvent sheet of the powder sample should be less than 6 ◦ ◦ show exactly the same type of distortion of the (i.e.theangledeviationof114 from120 ),inagree- hexagonalring,thiseffectislikelytobecausedbythe ment with the angle deviation from Meyer’s work interactionbetweensolventmoleculesandgraphene. [38],wheretheangledeviationfromthesheetis±5 . Nevertheless, further research needs to done to fully Likely, this effect is caused by the interaction of the confirmthisobservation. solventwithgraphene. Meyeretal[38]havesuggestedripplesinsuspen- ded graphene through TEM investigations, as con- firmed by other groups [39–41]. In particular, an 3.Conclusions average of 0.7 Å height fluctuation (i.e. ripple height normaltothesheet)wasfoundbyusingMonteCarlo Our work presents the first characterisation of simulation [42]. The bond length deviation is pre- solution-processed graphene in NMP and Cyrene dicted from 1.31 Å to 1.54 Å. This could include the by high energy x-ray scattering and related XPDF short double bond of 1.31 Å, a conjugated bond of analysis. The results show the disappearance of 1.42Å,anduptoalongsinglebondof1.54Å.Hence, long-range atomic correlations, confirming the pro- the three bonds of each carbon atom within the duction of 2D nanosheets and that the hexagonal hexagoncouldbedifferent.Ourresultsshowringdis- atomicstructureisstronglydistortedwhengraphene tortionscompatiblewiththeproposedmodelforsus- is suspended in a medium. In particular, the first pendedgraphene[39];astheC1–C3atomicdistance C–C distance slightly increases, while the second ofsolution-processedgraphenebecomesshorterthan C–Cdistancedecreases,resultinginadistortionthat the bulk graphite. However, it needs to be noted couldbecompatiblewithripplesformationobserved that TEM measurements were done on the suspen- in suspended graphene, likely to be caused by the dedgraphenewhileourmeasurementswereobtained interaction of solvent molecules with the graphene forgrapheneinsolution.Theincreasingbondlengths nanosheets.Ourresultsdemonstratethepotentialof differences indicate that the graphene’s increased XPDFasapowerfultooltocharacterise2Dmaterials density is not perfectly planar when dispersed in a andacquirequantitativestructuralinformationwith solvent. The bonding angle between C1–C2–C3 is atomicresolution. 6 2DMater.10(2023)015006 ZYanetal 4.Methods were used for the measurements. The x-ray pho- toelectron spectroscopy (XPS) measurements were 4.1.Materials performed using the K-Alpha x-ray Photoelectron High purity graphite foil (99.8% metal basis), Spectrometer (XPS) System from Thermo Scientific. graphite rod (99.99% metal basis) and ammonium The photon source was a monochromatized Al Kα sulfate ((NH ) SO , 98+%) were purchased line (hν = 1486.6 eV). The spectra were acquired 4 2 4 from Alfa Aesar. Anhydrous dimethyl sulfoxide usingaspotsizeof300 µmandconstantpassenergy (DMSO) (99.9%), graphite flakes (100+ mesh), (150 eV for survey and 20 eV for high resolution 1-Pyrenesulfonic acid sodium salt (PS1) and iso- spectra). A flood gun with combined electrons and propyl alcohol (IPA) were purchased from Sigma- lowenergyArionsisusedduringthemeasurements. Aldrich. Isomolded graphite (>99.95%) rods were HRTEM images were acquired on a JEOL 2100-F purchased from GraphiteStore. Cesium perchlorate microscope with a field-emission gun operated at (99%) was obtained from Fisher Scientific. The nat- 200 kV accelerating voltage providing direct images ural kish graphite was bought from Graphexel Ltd. of the atomic structure. A HAADF detector and an Allthechemicalsandmaterialswereusedasreceived. Oxford high solid-angle silicon drift detector x-ray energy dispersive spectrometer (EDS) system was 4.2.Exfoliation usedforchemicalelementalanalysis. The LPE graphene in NMP was prepared by adding 4.4.X-rayscattering 300 mg of graphite flakes into 100 ml of NMP, fol- lowed by sonicating the mixture at 600 W using The synchrotron x-ray scattering experiments were conductedinI15–1beamline,DiamondLightSource, Hilsonic bath sonicator for 5 d. Afterwards, the dis- persion was centrifuged using a Sigma 1–14k refri- U.K. The monochromatic x-ray beam with 76.7 keV (wavelength of 0.161669 Å) was employed [43]. A gerated centrifuge at 903 g for 20 min to remove un-exfoliatedgraphite.Toobtainhighlyconcentrated 2D Perkin Elmer XRD detector with active area of 2, 409.6 × 409.6 mm and pixel size of 100 µm was graphene, the dispersion was further centrifuged at 16600 g for 1 h., followed by re-dispersing the sed- applied close tosample to providelarge Q rangeand high-quality scattering data. Here Q = (4π sinθ) ⁄λ, imented graphene into a small volume of NMP. The ECispreparedbythecathodicelectrochemicalexfoli- where λ is the wavelength and 2θ is the angle between incident and scattered x-rays. The collec- ationasdescribedinourpreviouswork[12].Briefly, a pellet of natural graphite is used as a cathode, and teddiffractionintensitydataisprocessedbysoftware GudrunX [44] which subtracts the self-scattering Pt mesh was used as the anode. The electrolyte was 1M lithium chloride (Sigma Aldrich, 99.9%), and intensity, Compton scattering and multiple scatter- ing, etc. Then, the total scattering structure factor, Triethylamine hydrochloride in dimethyl sulfoxide. The exfoliation products were washed to remove the S(Q)andXPDF,G(r)areobtainedas; electrolyte with water and ethanol until the pH was I(Q) neutral,andtheproductswereseparatedbyfiltration S (Q) = (1) usingAnodiscaluminamembraneswith100nmpore size and then dried at 200 C under Ar atmosphere. max Thedrypowderwasthendispersedinasmallamount G(r) = Q[S(Q)−1]sin(Qr)dQ (2) ofNMA,asintheLPEsamples. π min 4.3.Materialscharacterization where I(Q) is the collected and processed diffraction UV–Vis spectroscopy of the graphene dispersions intensity. The coherent single-scattering intensity is were measured by using a PerkinElmer I-900 UV– desired; b is the element scattering amplitude (f is Vis–NIR spectrometer. A Bruker Atomic Froce usedforx-rayscattering);<…>denotesanaveraging Microscope (MultiMode 8) in Peak Force Tapping process.Fordetailedtheory,[21,45]isreferred. mode, equipped with ScanAsyst-Air tips is used to determine the lateral size and thickness distribution Dataavailabilitystatement of the graphene flakes. The samples were prepared by drop casting the dispersion on a clean silicon All data that support the findings of this study are substrate; several hundreds of individual flakes were included within the article (and any supplementary selected, after complete solvent evaporation, for lat- files). eral size and thickness analysis. The same sample preparationhasbeenusedforRamanmeasurements. Acknowledgments Raman measurements were performed using a Ren- ishaw Invia Raman spectrometer equipped with a We acknowledged Diamond Light Source for grant- 514.5 nm excitation line with 1 mW laser power. ing beamtime at I15-1 (CY24816). W M acknow- 100× NA0.85 objective lens, giving a spatial resol- ledges the funding from EPSRC (UK) Grant −1 ution of ∼500 nm, and 2400 grooves nm grating EP/P02680X/1. The work of M J G G is supported 7 2DMater.10(2023)015006 ZYanetal bytheXuntadeGalicia(Spain)PostdoctoralFellow- [8] TorrisiFetal2012Inkjet-printedgrapheneelectronicsACS Nano62992–3006 ship with reference ED481B-2019-015. C C and K P [9] NicolosiV,ChhowallaM,KanatzidisMG,StranoMSand acknowledgetheGrapheneFlagshipCore3(Contract ColemanJN2013Liquidexfoliationoflayeredmaterials No. 881603) and the ERC Project PEP2D (Contract Science3401226419 No. 770047). O R acknowledges financial support [10] BackesC,HigginsTM,KellyA,BolandC,HarveyA, HanlonDandColemanJN2017Guidelinesforexfoliation, from the Lloyd’s Register Foundation. C D and J U characterizationandprocessingoflayeredmaterials acknowledge the Office of Science, Office of Basic producedbyliquidexfoliationChem.Mater.29243–55 Energy Sciences, of the U.S. Department of Energy [11] HernandezYetal2008High-yieldproductionofgraphene (ContractNo.DE-AC02-05CH11231). byliquid-phaseexfoliationofgraphiteNat.Nanotechnol. 3563–8 [12] AbdelkaderAM,KinlochIAandDryfeRAW2014 Authorcontributions Continuouselectrochemicalexfoliationofmicrometer-sized grapheneusingsynergisticionintercalationsandorganic solventsACSAppl.Mater.Interfaces61632–9 Z Y, A A and W M conceived the approach and [13] AbdelkaderAM2015Electrochemicalsynthesisofhighly designed the experiments. A A, W M, and TF car- corrugatedgraphenesheetsforhighperformance ried out the x-ray experiments while M G and K P supercapacitorsJ.Mater.Chem.A38519–25 prepared materials and carried out lab experiments. [14] ParvezK,WuZ-S,LiR,LiuX,GrafR,FengXandMüllenK 2014Exfoliationofgraphiteintographeneinaqueous O R, C D and J U performed the material character- solutionsofinorganicsaltsJ.Am.Chem.Soc.1366083–91 izations with AFM, HRTEM, EDS and XPS analysis. [15] CooperAJ,WilsonNR,KinlochIAandDryfeRAW2014 Z Y performed the analysis the data and drafted the Singlestageelectrochemicalexfoliationmethodforthe manuscriptwithinputsfromCC,AAandWM.All productionoffew-layergrapheneviaintercalationof tetraalkylammoniumcationsCarbon66340–50 authors discussed the results and contributed to the [16] YangY,LuF,ZhouZ,SongW,ChenQandJiX2013 manuscript. Electrochemicallycathodicexfoliationofgraphenesheetsin roomtemperatureionicliquidsN-butyl, ORCIDiDs methylpyrrolidiniumbis(trifluoromethylsulfonyl)imideand theirelectrochemicalpropertiesElectrochim.Acta1139–16 [17] WickPetal2014Classificationframeworkfor ZhengyuYan https://orcid.org/0000-0002-2543- graphene-basedmaterialsAngew.Chem.Int.Ed.537714–8 [18] KostarelosKandNovoselovKS2014Graphenedevicesfor MaríaJGGuimarey https://orcid.org/0000-0003- lifeNat.Nanotechnol.9744–5 [19] PollardAJetal2017Characterisationofthestructureof 0483-4136 graphene(availableat:https://eprintspublications.npl.co. KhaledParvez https://orcid.org/0000-0003-2851- uk/8654/) [20] ISO/TS21356–12021Nanotechnologies—Structural ChaochaoDun https://orcid.org/0000-0002- CharacterizationofGraphene—Part1:Graphenefrom PowdersandDispersions(InternationalOrganisationfor 3215-6478 Standardization)(availableatwww.iso.org/standard/70757. OliverRead https://orcid.org/0000-0001-9771- html) [21] EgamiTandBillingeSJLEds2012UnderneaththeBragg JeffreyJUrban https://orcid.org/0000-0003-4909- PeaksStructuralAnalysisofComplexMaterialsvol 16(Oxford:Pergamon) [22] BilderbackDH,ElleaumePandWeckertE2005Reviewof CinziaCasiraghi https://orcid.org/0000-0001- thirdandnextgenerationsynchrotronlightsourcesJ.Phys. 7185-0377 B:At.Mol.Opt.Phys.38S773–97 WajiraMirihanage https://orcid.org/0000-0002- [23] HastingsJB,ThomlinsonWandCoxDE1984Synchrotron x-raypowderdiffractionJ.Appl.Crystallogr.1785–95 9083-269X [24] WestneatMW,SochaJJandLeeW-K2008Advancesin biologicalstructure,function,andphysiologyusing References synchrotronx-rayimagingAnnu.Rev.Physiol.70119–42 [25] SmithMA,FoleyHCandLoboRF2004Asimplemodel [1] BonaccorsoF,BartolottaA,ColemanJNandBackesC2016 describesthePDFofanon-graphitisingcarbonCarbon 2D-crystal-basedfunctionalinksAdv.Mater.286136–66 422041–8 [2] BackesCetal2020Productionandprocessingofgraphene [26] PetkovV,DifrancescoRG,BillingeSJL,AcharyaMand andrelatedmaterials2DMater.7022001 FoleyHC1999LocalstructureofnanoporouscarbonsPhil. [3] CaiX,LuoY,LiuBandChengH-M2018Preparationof2D Mag.B791519–30 materialdispersionsandtheirapplicationsChem.Soc.Rev. [27] WoznicaN,HawelekL,FischerHE,BobrinetskiyIand 476224–66 BurianA2015Theatomicscalestructureofgraphene [4] CiesielskiAandSamor`ıP2014Grapheneviasonication powderstudiedbyneutronandx-raydiffractionJ.Appl. assistedliquid-phaseexfoliationChem.Soc.Rev.43381–98 Crystallogr.481429–36 [5] KellyAGetal2017All-printedthin-filmtransistorsfrom [28] Wo´znicaN,HawelekL,DuberS,FischerHE,HonkimäkiV, networksofliquid-exfoliatednanosheetsScience35669–73 PawlytaM,BulouAandBurianA2017Theatomicscale [6] SecorEBandHersamMC2015Emergingcarbonand structureofsaccharose-basedcarbonsPhil.Mag. post-carbonnanomaterialinksforprintedelectronicsJ. 971675–97 Phys.Chem.Lett.6620–6 [29] SalavagioneHJ,SherwoodJ,DeBruynM,BudarinVL, [7] McManusDetal2017Water-basedandbiocompatible2D EllisGJ,ClarkJHandShuttleworthPS2017Identification crystalinksforall-inkjet-printedheterostructuresNat. ofhighperformancesolventsforthesustainableprocessing Nanotechnol.12343–50 ofgrapheneGreenChem.192550–60 8 2DMater.10(2023)015006 ZYanetal [30] ColemanJN2013Liquidexfoliationofdefect-freegraphene [38] MeyerJC,GeimAK,KatsnelsonMI,NovoselovKS, Acc.Chem.Res.4614–22 BoothTJandRothS2007Thestructureofsuspended [31] WangJ,MangaKK,BaoQandLohKP2011High-yield graphenesheetsNature44660–63 synthesisoffew-layergrapheneflakesthrough [39] AmiriA,NaraghiM,AhmadiG,SoleymanihaMand electrochemicalexpansionofgraphiteinpropylene ShanbediM2018Areviewonliquid-phaseexfoliationfor carbonateelectrolyteJ.Am.Chem.Soc.1338888–91 scalableproductionofpuregraphene,wrinkled,crumpled [32] ParvezK,WorsleyR,AlievaA,FeltenAandCasiraghiC andfunctionalizedgrapheneandchallengesFlatChem 2019Water-basedandinkjetprintableinksmadeby 840–71 electrochemicallyexfoliatedgrapheneCarbon149213–21 [40] LiZ,KinlochIA,YoungRJ,NovoselovKS, [33] NagyteVetal2020Ramanfingerprintsofgraphene AnagnostopoulosG,PartheniosJ,GaliotisC,PapagelisK, producedbyanodicelectrochemicalexfoliationNanoLett. LuC-YandBritnellL2015Deformationofwrinkled 203411–9 grapheneACSNano93917–25 [34] CasiraghiC,HartschuhA,QianH,PiscanecS,GeorgiC, [41] LiuN,PanZ,FuL,ZhangC,DaiBandLiuZ2011The FasoliA,NovoselovKS,BaskoDMandFerrariAC2009 originofwrinklesontransferredgrapheneNanoRes. RamanspectroscopyofgrapheneedgesNanoLett.91433–41 4996–1004 [35] ShangNG,PapakonstantinouP,SharmaS,LubarskyG, [42] FasolinoA,LosJHandKatsnelsonMI2007Intrinsic LiM,McNeillDW,QuinnAJ,ZhouWandBlackleyR2012 ripplesingrapheneNat.Mater.6858–61 Controllableselectiveexfoliationofhigh-qualitygraphene [43] ChaterPA,KeebleD,WharmbyM,SpainT,FilikJand nanosheetsandnanodotsbyionicliquidassistedgrinding WilhelmH2017TheautomatedXPDFbeamlineatdiamond Chem.Commun.481877 lightsourceActaCrystallogr.A73C69 [36] PetkovV,RenY,KabekkoduSandMurphyD2013Atomic [44] KeenDA2001Acomparisonofvariouscommonlyused pairdistributionfunctionsanalysisofdisorderedlow-Z correlationfunctionsfordescribingtotalscatteringJ.Appl. materialsPhys.Chem.Chem.Phys.158544 Crystallogr.34172–7 [37] BurianA,DoreJCandJurkiewiczK2019Structuralstudies [45] BillingeSJL2019Theriseofthex-rayatomicpair ofcarbonsbyneutronandx-rayscatteringRep.Prog.Phys. distributionfunctionmethod:aseriesoffortunateevents 82016501 Phil.Trans.R.Soc.A37720180413

Journal

2D MaterialsIOP Publishing

Published: Jan 1, 2023

Keywords: solution-processed graphene; x-ray pair distribution function; atomic structure

There are no references for this article.