Letter pubs.acs.org/ac
Characterization of Graphene-Nanoplatelets Structure via Thermogravimetry Michael Shtein,*,† Ilan Pri-Bar,*,‡ Maxim Varenik,‡ and Oren Regev*,†,‡ †
Ilse Katz Institute for Nanoscale Science and Technology and ‡Department of Chemical Engineering, Ben-Gurion University of the Negev, Beer-Sheva 8410501, Israel S Supporting Information *
ABSTRACT: The rapid increase in graphene-based applications has been accompanied by novel top-down manufacturing methods for graphene and its derivatives (e.g., graphene nanoplatelets (GnPs)). The characterization of the bulk properties of these materials by imaging and surface techniques (e.g., electron microscopy and Raman spectroscopy) is only possible through laborious and time-consuming statistical analysis, which precludes simple and efficient quality control during GnP production. We report that thermogravimetry (TG) may be utilized, beyond its conventional applications (e.g., quantification of impurities or surfactants, or labile functional groups) to characterize bulk GnP properties. We characterize the structural parameters of GnP (i.e., defect density, mean lateral dimension, and polydispersity) by imaging and surface techniques, on one hand, and by a systematic TG, on the other. The combined data demonstrate that the combustion temperature of commercially available and laboratoryprepared GnPs is correlated with their mean lateral dimension and defect density, while the combustion temperature range is proportional to their polydispersity index. Mapping all these parameters allows one to evaluate the GnPs’ structure following a simple thermogravimetric experiment (without necessitating further statistical analysis). Finally, TG is also used to detect and quantify different GnP constituents in powder and to conduct rapid quality-control tests during GnP production.
T
consuming statistical analysis,7 which currently precludes effective quality control during commercialized GnP production. Thermogravimetry (TG) is a common technique available in many research facilities, which offers rapid and simple characterization of bulk material properties.16,17 In regard to graphene production, it is mostly used to determine the amount of impurities (i.e, water, amorphous carbon, metals), labile functional groups,18−20 or leftover traces of surfactants (following a surfactant-assisted exfoliation21−23) in nanocarbon powders. Only a few researchers have applied TG to the characterization of various carbonaceous species24,25 and studied their decomposition temperatures26 or their thermal stability performance.27 In this study, we employ TG, beyond its conventional analysis, to characterize bulk GnP properties. To demonstrate this, we statistically characterize the structural parameters of various commercially available and laboratory-prepared GnPs by imaging and surface techniques and show that they correlate very well with thermal parameters extracted from GnP thermogravimetric curves.
he interest in graphene has been growing rapidly over the past several years due to its unique mechanical,1 thermal,2 and electrical3 properties. These characteristics are advantageous in diverse applications, ranging from solar cells4 to composite materials.5 The wealth of applications calls for an increased demand for graphene and its derivatives (e.g., graphene nanoplatelets (GnPs), 3−10 stacks of graphene). Recently, a few promising scalable top-down manufacturing methods have emerged, in which graphite is the raw material. Among these are sonication,6 high shear mixing,7 thermal intercalation,8 and ballmilling,9 which result in either graphene or GnP powder, or both, with various lateral dimension, thickness, polydispersity, defect density, and impurities. Utilization of graphene in commercial applications requires an effective quality control of these properties. Thus, simple and rapid characterization techniques are necessary. Graphene has often been characterized by imaging techniques, such as scanning electron microscopy (SEM),10 transmission electron microscopy11 (TEM), and atomic force microscopy (AFM), to provide its lateral dimension and thickness. Raman spectroscopy (RS) is used to evaluate the defect concentration (sp3 hybridization) and the number of graphene layers in GnPs.12−14 X-ray photoelectron spectroscopy (XPS) provides surface information on the chemical species (e.g., sp2 or sp3 carbons).15 These techniques characterize the properties of either a few particles (EM, RS) or sample surfaces (XPS) but not the bulk. The latter is only possible through laborious and time© XXXX American Chemical Society
■
EXPERIMENTAL SECTION Materials. Commercial grade C500, C750, Mx, and Hx GnPs (x is the mean lateral dimension (MLD) and equal to 5, 15, or 25 Received: January 18, 2015 Accepted: March 22, 2015
A
DOI: 10.1021/acs.analchem.5b00228 Anal. Chem. XXXX, XXX, XXX−XXX
Letter
Analytical Chemistry μm) were purchased from XG-Sciences; grade 3 (G3) and grade 4 (G4) GnPs from Cheaptubes, and Graphite flakes (CAS 7782-425) from Sigma-Aldrich. All chemicals were used as received. Laboratory-Prepared GnPs (Lx Grade). GnPs were prepared by recently published procedures, namely, exfoliating graphite flakes by tip sonication21 or by ball mill grinding.28 By changing the sonication power or milling speed, we prepared GnPs (Lx) with various MLD (x = 0.2−0.7 μm) and defect densities (55−80% sp2). Thermogravimetry (TG). The thermogravimetric curves of GnP powders were recorded by a Mettler Toledo analyzer with a Stare software system (TGA/STDA85). The TG was carried out under air flow (50 mL/min). The samples (4−6 mg in 70 μL alumina crucibles) were heated from 40 to 500 °C at a rate of 10 °C/min, and from 500 to 1000 °C at a 5 °C/min rate. Decreasing the heating rate (to 5 °C/min) was essential, since it significantly enhanced the detection of thermal events in the relevant temperature range (500 to 1000 °C). Few thermal parameters were extracted: The mean combustion temperature (CT, star in Figure 1d) was defined as the temperature at which the
Electron Microscopy (EM). High-resolution cold FEG SEM (JSM-7400F (JEOL)) or TEM (FEI TECNAI T12) were used to determine the GnP lateral dimensions. The mean lateral dimension (MLD) and polydispersity index (PDI = standard deviation/MLD) calculations are based on at least 100 particles. The PDI values range from 0 to 1; a higher value indicates a less homogeneous GnP size distribution, i.e., a higher polydispersity. Atomic Force Microscopy (AFM). GnP thicknesses were determined by Dimension 3100 SPM in tapping mode using Veeco RTESP silicon tips. GnP aqueous dispersion was spincoated on SiO2 wafers and allowed to dry by evaporation at ambient temperature for 24 h before measurement. Raman Spectroscopy (RS). The number of graphene layers and relative defects12 in the GnP were determined by Jobin-Yvon HR LabRam micro-Raman operated at 514 nm on a quartz slide. The Raman spectrum is characterized by three major bands: (i) the intensity of D band (∼1350 cm−1) is due to the first-order phonons and indicates the in-plane and edges defect density; (ii) the G band (∼1580 cm−1) relates to the zone center Ramanallowed band; and (iii) the 2D band relates to the second-order phonons, and its shift indicates the relative GnP thickness (∼2727 cm−1 for graphite and 900 °C, Table S1, Supporting Information). The in-plane defects become vacancies at higher temperatures;34 hence, GnPs with lower values of sp2% are characterized by lower CT, as well. The mapping (Figure 2c) of all our experimental data allows one to evaluate the GnP structural parameters following a simple thermogravimetric experiment (without further statistical analysis). GnP Polydispersity. Since structural parameters dictate the GnP properties, highly polydispersed materials may cause unpredictable or nonideal behavior.35 Thus, GnPs with narrow polydispersity are essential for future applications. Presently, the determination of GnP polydispersity requires laborious manual statistical analysis of a handful of TEM micrographs7 (vide supra). TG can easily provide an estimation of bulk GnP polydispersity by measuring the temperature range, in which the combustion process takes place (i.e., the CTR, red dots in Figure 1d). We found a linear correlation (CTR = (306 ± 8) × PDI, Figure 3a) between the CTR (extracted from TG) and the PDI (extracted from EM statistical analysis). On the basis of this correlation, a monodispersed powder (PDI → 0) has a CTR of 0 °C, while a combustion process of highly polydispersed powder (PDI → 1) stretches over a wide temperature range (CTR (306 ± 8 °C)). For instance, the combustion process of particles with narrow polydispersity (e.g., L0.2 with PDI = 0.26) takes place in a narrow range: CTRm = 82 ± 2 °C (measured via TG) or CTRc = 80 ± 8 °C (calculated from the above linear correlation). However, the CTR of widely polydispersed GnPs (e.g., H25 with PDI = 0.59) increases significantly (CTRm = 185 ± 2 °C or CTRc = 181 ± 8 °C). Here, the smaller GnPs start combust at lower temperature, while the larger ones do so at significantly higher temperatures (Section 1 of the Supporting Information).
Raman spectroscopy (RS) reveals that the GnP spectra are characteristic of graphene (Figure 1b). In particular, they all show identical 2D band characteristics, typical of a few layer graphene;7,12,32 namely, the thicknesses of all GnPs in this study are equal (within the experimental error). This observation was confirmed by AFM (Figure S1d). The D-to-G-peak intensity ratio is below 0.3, suggesting a low defect density of GnPs.33 A typical deconvoluted XPS C 1s spectra shows (Figure 1c) that the intensity of sp2 (CC bonds) is much higher than the sp3 (C−C and C−O bonds found in edges and defects) manifesting low defect density. The variation in sp2 densities for various GnP grades is presented in Figure 2a,b (right-ordinate) and in Table S1, Supporting Information. The characterization results of commercially available GnP (Hx, Mx, Cx, and Gx, Table S1, Supporting Information) agree with the specifications provided by the vendors of those GnPs. MLD-sp2-CT Correlations. A typical thermogravimetric curve of GnP powder is presented in Figure 1d. Its first derivative (derivative thermogravimetry (DTG)) facilitates the detection of thermal events. Commonly, the first thermal event (at ∼200−560 °C, dotted blue step) is due to the combustion of amorphous carbon,24,25 while the second event results from GnP combustion (∼594−723 °C, dashed black step). The remaining residues (15 wt %) consist of SiO2, Al2O3, and Fe2O3 impurities, originating from the raw graphite material, confirmed by the XPS analysis. Our systematic analysis of numerous GnP thermogravimetric curves indicates that CT, MLD, and sp2% are strongly dependent; CT values are very sensitive to mild changes in GnP MLD and defect density (100-sp2%), as demonstrated for laboratoryprepared GnPs (Figure 2a). For example, an increase in the MLD (from 0.2 to 0.7 μm) and the amount of sp2 bonds (55% → 81%) resulted in an increased CT (647 °C → 687 °C). Furthermore, we found that MLD is proportional to the CT (Figure 2b, MLD = 0.336 × CT − 218), and sp2% scales with CT (Figure S2, Supporting Information). A similar behavior was recorded for all samples (Table S1, Supporting Information). The explanation of C
DOI: 10.1021/acs.analchem.5b00228 Anal. Chem. XXXX, XXX, XXX−XXX
Letter
Analytical Chemistry
Figure 3. (a) The combustion temperature range (CTR, extracted from TG) vs the polydispersity index (PDI, extracted from TEM) of selected GnPs and their linear fit (red dashed line); (b) normalized thermogravimetric curves; (c) DTG curve of the GnP mixtureA (L0.5/M15) and its pure constituents (L0.5 and M15). Inset: the log-normal deconvolution of the L0.5/M15 DTG curve shows an excellent match to maximum combustion rate temperatures of the pure GnP species. The blue dotted curve denotes a fit to the experimental data (black line). The areas under the deconvoluted curves are indicative of weight fraction ratio of the GnPs constituents.
■
Quality Control via TG. We further investigated the effectiveness of our TG-based method in detection and quantification of nanocarbon constituents and its utilization for products quality control. In demonstration, mixtures containing particles with significantly different MLDs were prepared (Section 3 of the Supporting Information). Such samples could indicate a defective product due to an uncontrolled production process. Figure 3b,c shows TG and DTG curves of pure components (L0.5 and M15) and of their mixtureA (L0.5/M15, 1:1 weight ratio). In the mixtureA DTG curve (Figure 3b, full black line), two thermal events are observed and assigned to the combustion of the different carbonaceous constituents; the first thermal event (400−500 °C) is assigned to the combustion of an amorphous carbon and the second (575−750 °C) to GnP combustion. An in-depth analysis of the DTG curve of mixtureA (Figure 3c, full black line) reveals that the second thermal event consists of two separate events. The discretization of these two events (Figure 3c inset) is accomplished by log-normal deconvolution (Section 1 of the Supporting Information); the event at 675 °C is assigned to the combustion of L0.5 GnP of smaller lateral dimension, while the event at 727 °C is assigned to the M15 GnP, having significantly larger lateral dimensions. Remarkably, the temperatures, at which the pure constituents combustion rate is maximal (defined by minimum value of the DTG curve), are in excellent agreement with the values derived from the deconvoluted curves (680 °C vs 675 °C for L0.5 and 729 °C vs 727 °C for D15), while the integrals’ ratio (1:1, highlighted areas in Figure 3c inset) is equal to the weight fraction ratio of the constituents (1:1). These results validate that TG may be utilized for detecting and quantifying nanocarbon constituents. To demonstrate that the proposed analysis may also serve as an efficient quality control technique, a defective product was analyzed (mixtureB, Section 3 of the Supporting Information): the DTG curve deconvolution of the main thermal event shows that it consists of two separate events (Figure S3, Supporting Information). The first thermal event, at 740 °C, is assigned to the combustion of the exfoliated GnPs, while the second belongs to the defective unexfoliated graphite fraction (CT > 900 °C). The ratio of the areas under the deconvoluted curves (0.3:1) is indicative of the weight fraction ratio of the GNP/graphite constituents. The proposed method allows detection of a small defective fraction (5 wt %, Figure S4, Supporting Information).
CONCLUSIONS We showed that TG is a straightforward and effective diagnostic tool for GnP products. Advanced TG of numerous GnPs provide information on their structure (dimension, defect density, polydispersity) and are capable of detecting nanocarbon types and of determining their composition in mixtures. The map in Figure 2c, constructed by combined thermogravimetry−imagingsurface techniques, offers a simple and rapid bulk property characterization, which may be applied in GnP production plants.
■
ASSOCIATED CONTENT
S Supporting Information *
DTG curve deconvolution, typical characterization of as-received commercial materials, CT-sp2% dependence, GnP mixtures for constituent detection, and quantification experiments. This material is available free of charge via the Internet at http:// pubs.acs.org/.
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail: [email protected]. *E-mail: [email protected]. Tel.: +972 86472763. *E-mail: [email protected]. Tel.: +972 86472145. Author Contributions
The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes
The authors declare no competing financial interest.
■
REFERENCES
(1) Frank, I. W.; Tanenbaum, D. M.; van der Zande, A. M.; McEuen, P. L. J. Vac. Sci. Technol., B: Nanotechnol. Microelectron.: Mater., Process., Meas., Phenom. 2007, 25, 2558−2561. (2) Balandin, A. A.; Ghosh, S.; Bao, W.; Calizo, I.; Teweldebrhan, D.; Miao, F.; Lau, C. N. Nano Lett. 2008, 8, 902−907. (3) Castro Neto, A. H.; Guinea, F.; Peres, N. M. R.; Novoselov, K. S.; Geim, A. K. Rev. Mod. Phys. 2009, 81, 109−162. (4) Kavan, L.; Yum, J. H.; Grätzel, M. ACS Nano 2010, 5, 165−172. (5) Huang, X.; Qi, X.; Boey, F.; Zhang, H. Chem. Soc. Rev. 2012, 41, 666−686. (6) Ciesielski, A.; Samori, P. Chem. Soc. Rev. 2014, 43, 381−398. (7) Paton, K. R.; Varrla, E.; Backes, C.; Smith, R. J.; Khan, U.; O’Neill, A.; Boland, C.; Lotya, M.; Istrate, O. M.; King, P.; Higgins, T.; Barwich, S.; May, P.; Puczkarski, P.; Ahmed, I.; Moebius, M.; Pettersson, H.; Long, E.;
D
DOI: 10.1021/acs.analchem.5b00228 Anal. Chem. XXXX, XXX, XXX−XXX
Letter
Analytical Chemistry Coelho, J.; O’Brien, S. E.; McGuire, E. K.; Sanchez, B. M.; Duesberg, G. S.; McEvoy, N.; Pennycook, T. J.; Downing, C.; Crossley, A.; Nicolosi, V.; Coleman, J. N. Nat. Mater. 2014, 13, 624−630. (8) Lee, S.; Cho, D.; Drzal, L. T. J. Mater. Sci. 2005, 40, 231−234. (9) Jeon, I. Y.; Shin, Y. R.; Sohn, G. J.; Choi, H. J.; Bae, S. Y.; Mahmood, J.; Jung, S. M.; Seo, J. M.; Kim, M. J.; Wook Chang, D.; Dai, L.; Baek, J. B. Proc. Natl. Acad. Sci. 2012, 109, 5588−5593. (10) Xie, J.; Spallas, J. P. Different contrast mechanisms in SEM imaging of graphene, Technical Report, Agilent Technologies; http://www.toyo.co. jp/file/pdf/spm/files/agi/5991-0782EN.pdf, 2012. (11) Wang, G.; Yang, J.; Park, J.; Gou, X.; Wang, B.; Liu, H.; Yao, J. J. Phys. Chem. C 2008, 112, 8192−8195. (12) Ferrari, A. C.; Meyer, J. C.; Scardaci, V.; Casiraghi, C.; Lazzeri, M.; Mauri, F.; Piscanec, S.; Jiang, D.; Novoselov, K. S.; Roth, S.; Geim, A. K. Phys. Rev. Lett. 2006, 97, 187401. (13) Chae, S. J.; Gunes, F.; Kim, K. K.; Kim, E. S.; Han, G. H.; Kim, S. M.; Shin, H. J.; Yoon, S. M.; Choi, J. Y.; Park, M. H.; Yang, C. W.; Pribat, D.; Lee, Y. H. Adv. Mater. 2009, 21, 2328−2333. (14) Reina, A.; Jia, X. T.; Ho, J.; Nezich, D.; Son, H. B.; Bulovic, V.; Dresselhaus, M. S.; Kong, J. Nano Lett. 2009, 9, 30−35. (15) Barron, A. R.; Flood, D. J. XPS of Carbon Nanomaterials. In OpenStax CNX, 2013; http://cnx.org/contents/[email protected]:10/Carbon_Nanotubes. (16) Gabbott, P. Principles and Applications of Thermal Analysis; Wiley: New York, 2008. (17) Haines, P. J. Principles of Thermal Analysis and Calorimetry; Royal Society of Chemistry: Cambridge, U.K., 2002. (18) Loryuenyong, V.; Totepvimarn, K.; Eimburanapravat, P.; Boonchompoo, W.; Buasri, A. Adv. Mater. Sci. Eng. 2013, 2013, 5. (19) Kishore, M.; Srimathi, P.; Kumar, S.; Addepalli, S.; Swaminathan, S.; Tilak, V.; Colborn, R. Bull. Mater. Sci. 2013, 36, 667−672. (20) Chen, W.; Yan, L. Nanoscale 2010, 2, 559−563. (21) Chabot, V.; Kim, B.; Sloper, B.; Tzoganakis, C.; Yu, A. Sci. Rep. 2013, 3, 1378. (22) Parviz, D.; Das, S.; Ahmed, H. S. T.; Irin, F.; Bhattacharia, S.; Green, M. J. ACS Nano 2012, 6, 8857−8867. (23) Buzaglo, M.; Shtein, M.; Kober, S.; Lovrincic, R.; Vilan, A.; Regev, O. Phys. Chem. Chem. Phys. 2013, 15, 4428−4435. (24) Gallego, J.; Batiot-Dupeyat, C.; Mondragón, F. J. Therm. Anal. Calorim. 2013, 114, 597−602. (25) Cataldo, F. Fullerenes, Nanotubes, Carbon Nanostruct. 2002, 10, 293−311. (26) Sahle-Demessie, E.; Zhao, A.; Salamon, A. W. A Study of Aged Carbon Nanotubes by Thermogravimetric Analysis; PerkinElmer, Inc.: Shelton, CT, 2011. (27) Mahajan, A.; Kingon, A.; Kukovecz, Á .; Konya, Z.; Vilarinho, P. M. Mater. Lett. 2013, 90, 165−168. (28) León, V.; Rodriguez, A. M.; Prieto, P.; Prato, M.; Vázquez, E. ACS Nano 2013, 8, 563−571. (29) Schawe, J. E. K. Elastomers; Mettler-Toledo: Columbus, OH, 2002; Vol. 2. (30) Memon, N. K.; Tse, S. D.; Al-Sharab, J. F.; Yamaguchi, H.; Goncalves, A. M. B.; Kear, B. H.; Jaluria, Y.; Andrei, E. Y.; Chhowalla, M. Carbon 2011, 49, 5064−5070. (31) Loomis, J.; Fan, X.; Khosravi, F.; Xu, P.; Fletcher, M.; Cohn, R. W.; Panchapakesan, B. Sci. Rep. 2013, 3, 1900. (32) Duhee Yoon, H. M. A. H. C. J. Korean Phys. Soc. 2009, 55, 1299− 1303. (33) Cançado, L. G.; Jorio, A.; Ferreira, E. H. M.; Stavale, F.; Achete, C. A.; Capaz, R. B.; Moutinho, M. V. O.; Lombardo, A.; Kulmala, T. S.; Ferrari, A. C. Nano Lett. 2011, 11, 3190−3196. (34) Nan, H. Y.; Ni, Z. H.; Wang, J.; Zafar, Z.; Shi, Z. X.; Wang, Y. Y. J. Raman Spectrosc. 2013, 44, 1018−1021. (35) Green, A. A.; Hersam, M. C. J. Phys. Chem. Lett. 2009, 1, 544−549. (36) Seo, J. W. T.; Green, A. A.; Antaris, A. L.; Hersam, M. C. J. Phys. Chem. Lett. 2011, 2, 1004−1008.
E
DOI: 10.1021/acs.analchem.5b00228 Anal. Chem. XXXX, XXX, XXX−XXX
Characterization of Graphene-Nanoplatelets Structure via Thermogravimetry Michael Shtein,*,† Ilan Pri-Bar,*,‡ Maxim Varenik,‡ and Oren Regev*,†,‡ †
Ilse Katz Institute for Nanoscale Science and Technology and ‡Department of Chemical Engineering, Ben-Gurion University of the Negev, Beer-Sheva 8410501, Israel S Supporting Information *
ABSTRACT: The rapid increase in graphene-based applications has been accompanied by novel top-down manufacturing methods for graphene and its derivatives (e.g., graphene nanoplatelets (GnPs)). The characterization of the bulk properties of these materials by imaging and surface techniques (e.g., electron microscopy and Raman spectroscopy) is only possible through laborious and time-consuming statistical analysis, which precludes simple and efficient quality control during GnP production. We report that thermogravimetry (TG) may be utilized, beyond its conventional applications (e.g., quantification of impurities or surfactants, or labile functional groups) to characterize bulk GnP properties. We characterize the structural parameters of GnP (i.e., defect density, mean lateral dimension, and polydispersity) by imaging and surface techniques, on one hand, and by a systematic TG, on the other. The combined data demonstrate that the combustion temperature of commercially available and laboratoryprepared GnPs is correlated with their mean lateral dimension and defect density, while the combustion temperature range is proportional to their polydispersity index. Mapping all these parameters allows one to evaluate the GnPs’ structure following a simple thermogravimetric experiment (without necessitating further statistical analysis). Finally, TG is also used to detect and quantify different GnP constituents in powder and to conduct rapid quality-control tests during GnP production.
T
consuming statistical analysis,7 which currently precludes effective quality control during commercialized GnP production. Thermogravimetry (TG) is a common technique available in many research facilities, which offers rapid and simple characterization of bulk material properties.16,17 In regard to graphene production, it is mostly used to determine the amount of impurities (i.e, water, amorphous carbon, metals), labile functional groups,18−20 or leftover traces of surfactants (following a surfactant-assisted exfoliation21−23) in nanocarbon powders. Only a few researchers have applied TG to the characterization of various carbonaceous species24,25 and studied their decomposition temperatures26 or their thermal stability performance.27 In this study, we employ TG, beyond its conventional analysis, to characterize bulk GnP properties. To demonstrate this, we statistically characterize the structural parameters of various commercially available and laboratory-prepared GnPs by imaging and surface techniques and show that they correlate very well with thermal parameters extracted from GnP thermogravimetric curves.
he interest in graphene has been growing rapidly over the past several years due to its unique mechanical,1 thermal,2 and electrical3 properties. These characteristics are advantageous in diverse applications, ranging from solar cells4 to composite materials.5 The wealth of applications calls for an increased demand for graphene and its derivatives (e.g., graphene nanoplatelets (GnPs), 3−10 stacks of graphene). Recently, a few promising scalable top-down manufacturing methods have emerged, in which graphite is the raw material. Among these are sonication,6 high shear mixing,7 thermal intercalation,8 and ballmilling,9 which result in either graphene or GnP powder, or both, with various lateral dimension, thickness, polydispersity, defect density, and impurities. Utilization of graphene in commercial applications requires an effective quality control of these properties. Thus, simple and rapid characterization techniques are necessary. Graphene has often been characterized by imaging techniques, such as scanning electron microscopy (SEM),10 transmission electron microscopy11 (TEM), and atomic force microscopy (AFM), to provide its lateral dimension and thickness. Raman spectroscopy (RS) is used to evaluate the defect concentration (sp3 hybridization) and the number of graphene layers in GnPs.12−14 X-ray photoelectron spectroscopy (XPS) provides surface information on the chemical species (e.g., sp2 or sp3 carbons).15 These techniques characterize the properties of either a few particles (EM, RS) or sample surfaces (XPS) but not the bulk. The latter is only possible through laborious and time© XXXX American Chemical Society
■
EXPERIMENTAL SECTION Materials. Commercial grade C500, C750, Mx, and Hx GnPs (x is the mean lateral dimension (MLD) and equal to 5, 15, or 25 Received: January 18, 2015 Accepted: March 22, 2015
A
DOI: 10.1021/acs.analchem.5b00228 Anal. Chem. XXXX, XXX, XXX−XXX
Letter
Analytical Chemistry μm) were purchased from XG-Sciences; grade 3 (G3) and grade 4 (G4) GnPs from Cheaptubes, and Graphite flakes (CAS 7782-425) from Sigma-Aldrich. All chemicals were used as received. Laboratory-Prepared GnPs (Lx Grade). GnPs were prepared by recently published procedures, namely, exfoliating graphite flakes by tip sonication21 or by ball mill grinding.28 By changing the sonication power or milling speed, we prepared GnPs (Lx) with various MLD (x = 0.2−0.7 μm) and defect densities (55−80% sp2). Thermogravimetry (TG). The thermogravimetric curves of GnP powders were recorded by a Mettler Toledo analyzer with a Stare software system (TGA/STDA85). The TG was carried out under air flow (50 mL/min). The samples (4−6 mg in 70 μL alumina crucibles) were heated from 40 to 500 °C at a rate of 10 °C/min, and from 500 to 1000 °C at a 5 °C/min rate. Decreasing the heating rate (to 5 °C/min) was essential, since it significantly enhanced the detection of thermal events in the relevant temperature range (500 to 1000 °C). Few thermal parameters were extracted: The mean combustion temperature (CT, star in Figure 1d) was defined as the temperature at which the
Electron Microscopy (EM). High-resolution cold FEG SEM (JSM-7400F (JEOL)) or TEM (FEI TECNAI T12) were used to determine the GnP lateral dimensions. The mean lateral dimension (MLD) and polydispersity index (PDI = standard deviation/MLD) calculations are based on at least 100 particles. The PDI values range from 0 to 1; a higher value indicates a less homogeneous GnP size distribution, i.e., a higher polydispersity. Atomic Force Microscopy (AFM). GnP thicknesses were determined by Dimension 3100 SPM in tapping mode using Veeco RTESP silicon tips. GnP aqueous dispersion was spincoated on SiO2 wafers and allowed to dry by evaporation at ambient temperature for 24 h before measurement. Raman Spectroscopy (RS). The number of graphene layers and relative defects12 in the GnP were determined by Jobin-Yvon HR LabRam micro-Raman operated at 514 nm on a quartz slide. The Raman spectrum is characterized by three major bands: (i) the intensity of D band (∼1350 cm−1) is due to the first-order phonons and indicates the in-plane and edges defect density; (ii) the G band (∼1580 cm−1) relates to the zone center Ramanallowed band; and (iii) the 2D band relates to the second-order phonons, and its shift indicates the relative GnP thickness (∼2727 cm−1 for graphite and 900 °C, Table S1, Supporting Information). The in-plane defects become vacancies at higher temperatures;34 hence, GnPs with lower values of sp2% are characterized by lower CT, as well. The mapping (Figure 2c) of all our experimental data allows one to evaluate the GnP structural parameters following a simple thermogravimetric experiment (without further statistical analysis). GnP Polydispersity. Since structural parameters dictate the GnP properties, highly polydispersed materials may cause unpredictable or nonideal behavior.35 Thus, GnPs with narrow polydispersity are essential for future applications. Presently, the determination of GnP polydispersity requires laborious manual statistical analysis of a handful of TEM micrographs7 (vide supra). TG can easily provide an estimation of bulk GnP polydispersity by measuring the temperature range, in which the combustion process takes place (i.e., the CTR, red dots in Figure 1d). We found a linear correlation (CTR = (306 ± 8) × PDI, Figure 3a) between the CTR (extracted from TG) and the PDI (extracted from EM statistical analysis). On the basis of this correlation, a monodispersed powder (PDI → 0) has a CTR of 0 °C, while a combustion process of highly polydispersed powder (PDI → 1) stretches over a wide temperature range (CTR (306 ± 8 °C)). For instance, the combustion process of particles with narrow polydispersity (e.g., L0.2 with PDI = 0.26) takes place in a narrow range: CTRm = 82 ± 2 °C (measured via TG) or CTRc = 80 ± 8 °C (calculated from the above linear correlation). However, the CTR of widely polydispersed GnPs (e.g., H25 with PDI = 0.59) increases significantly (CTRm = 185 ± 2 °C or CTRc = 181 ± 8 °C). Here, the smaller GnPs start combust at lower temperature, while the larger ones do so at significantly higher temperatures (Section 1 of the Supporting Information).
Raman spectroscopy (RS) reveals that the GnP spectra are characteristic of graphene (Figure 1b). In particular, they all show identical 2D band characteristics, typical of a few layer graphene;7,12,32 namely, the thicknesses of all GnPs in this study are equal (within the experimental error). This observation was confirmed by AFM (Figure S1d). The D-to-G-peak intensity ratio is below 0.3, suggesting a low defect density of GnPs.33 A typical deconvoluted XPS C 1s spectra shows (Figure 1c) that the intensity of sp2 (CC bonds) is much higher than the sp3 (C−C and C−O bonds found in edges and defects) manifesting low defect density. The variation in sp2 densities for various GnP grades is presented in Figure 2a,b (right-ordinate) and in Table S1, Supporting Information. The characterization results of commercially available GnP (Hx, Mx, Cx, and Gx, Table S1, Supporting Information) agree with the specifications provided by the vendors of those GnPs. MLD-sp2-CT Correlations. A typical thermogravimetric curve of GnP powder is presented in Figure 1d. Its first derivative (derivative thermogravimetry (DTG)) facilitates the detection of thermal events. Commonly, the first thermal event (at ∼200−560 °C, dotted blue step) is due to the combustion of amorphous carbon,24,25 while the second event results from GnP combustion (∼594−723 °C, dashed black step). The remaining residues (15 wt %) consist of SiO2, Al2O3, and Fe2O3 impurities, originating from the raw graphite material, confirmed by the XPS analysis. Our systematic analysis of numerous GnP thermogravimetric curves indicates that CT, MLD, and sp2% are strongly dependent; CT values are very sensitive to mild changes in GnP MLD and defect density (100-sp2%), as demonstrated for laboratoryprepared GnPs (Figure 2a). For example, an increase in the MLD (from 0.2 to 0.7 μm) and the amount of sp2 bonds (55% → 81%) resulted in an increased CT (647 °C → 687 °C). Furthermore, we found that MLD is proportional to the CT (Figure 2b, MLD = 0.336 × CT − 218), and sp2% scales with CT (Figure S2, Supporting Information). A similar behavior was recorded for all samples (Table S1, Supporting Information). The explanation of C
DOI: 10.1021/acs.analchem.5b00228 Anal. Chem. XXXX, XXX, XXX−XXX
Letter
Analytical Chemistry
Figure 3. (a) The combustion temperature range (CTR, extracted from TG) vs the polydispersity index (PDI, extracted from TEM) of selected GnPs and their linear fit (red dashed line); (b) normalized thermogravimetric curves; (c) DTG curve of the GnP mixtureA (L0.5/M15) and its pure constituents (L0.5 and M15). Inset: the log-normal deconvolution of the L0.5/M15 DTG curve shows an excellent match to maximum combustion rate temperatures of the pure GnP species. The blue dotted curve denotes a fit to the experimental data (black line). The areas under the deconvoluted curves are indicative of weight fraction ratio of the GnPs constituents.
■
Quality Control via TG. We further investigated the effectiveness of our TG-based method in detection and quantification of nanocarbon constituents and its utilization for products quality control. In demonstration, mixtures containing particles with significantly different MLDs were prepared (Section 3 of the Supporting Information). Such samples could indicate a defective product due to an uncontrolled production process. Figure 3b,c shows TG and DTG curves of pure components (L0.5 and M15) and of their mixtureA (L0.5/M15, 1:1 weight ratio). In the mixtureA DTG curve (Figure 3b, full black line), two thermal events are observed and assigned to the combustion of the different carbonaceous constituents; the first thermal event (400−500 °C) is assigned to the combustion of an amorphous carbon and the second (575−750 °C) to GnP combustion. An in-depth analysis of the DTG curve of mixtureA (Figure 3c, full black line) reveals that the second thermal event consists of two separate events. The discretization of these two events (Figure 3c inset) is accomplished by log-normal deconvolution (Section 1 of the Supporting Information); the event at 675 °C is assigned to the combustion of L0.5 GnP of smaller lateral dimension, while the event at 727 °C is assigned to the M15 GnP, having significantly larger lateral dimensions. Remarkably, the temperatures, at which the pure constituents combustion rate is maximal (defined by minimum value of the DTG curve), are in excellent agreement with the values derived from the deconvoluted curves (680 °C vs 675 °C for L0.5 and 729 °C vs 727 °C for D15), while the integrals’ ratio (1:1, highlighted areas in Figure 3c inset) is equal to the weight fraction ratio of the constituents (1:1). These results validate that TG may be utilized for detecting and quantifying nanocarbon constituents. To demonstrate that the proposed analysis may also serve as an efficient quality control technique, a defective product was analyzed (mixtureB, Section 3 of the Supporting Information): the DTG curve deconvolution of the main thermal event shows that it consists of two separate events (Figure S3, Supporting Information). The first thermal event, at 740 °C, is assigned to the combustion of the exfoliated GnPs, while the second belongs to the defective unexfoliated graphite fraction (CT > 900 °C). The ratio of the areas under the deconvoluted curves (0.3:1) is indicative of the weight fraction ratio of the GNP/graphite constituents. The proposed method allows detection of a small defective fraction (5 wt %, Figure S4, Supporting Information).
CONCLUSIONS We showed that TG is a straightforward and effective diagnostic tool for GnP products. Advanced TG of numerous GnPs provide information on their structure (dimension, defect density, polydispersity) and are capable of detecting nanocarbon types and of determining their composition in mixtures. The map in Figure 2c, constructed by combined thermogravimetry−imagingsurface techniques, offers a simple and rapid bulk property characterization, which may be applied in GnP production plants.
■
ASSOCIATED CONTENT
S Supporting Information *
DTG curve deconvolution, typical characterization of as-received commercial materials, CT-sp2% dependence, GnP mixtures for constituent detection, and quantification experiments. This material is available free of charge via the Internet at http:// pubs.acs.org/.
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail: [email protected]. *E-mail: [email protected]. Tel.: +972 86472763. *E-mail: [email protected]. Tel.: +972 86472145. Author Contributions
The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes
The authors declare no competing financial interest.
■
REFERENCES
(1) Frank, I. W.; Tanenbaum, D. M.; van der Zande, A. M.; McEuen, P. L. J. Vac. Sci. Technol., B: Nanotechnol. Microelectron.: Mater., Process., Meas., Phenom. 2007, 25, 2558−2561. (2) Balandin, A. A.; Ghosh, S.; Bao, W.; Calizo, I.; Teweldebrhan, D.; Miao, F.; Lau, C. N. Nano Lett. 2008, 8, 902−907. (3) Castro Neto, A. H.; Guinea, F.; Peres, N. M. R.; Novoselov, K. S.; Geim, A. K. Rev. Mod. Phys. 2009, 81, 109−162. (4) Kavan, L.; Yum, J. H.; Grätzel, M. ACS Nano 2010, 5, 165−172. (5) Huang, X.; Qi, X.; Boey, F.; Zhang, H. Chem. Soc. Rev. 2012, 41, 666−686. (6) Ciesielski, A.; Samori, P. Chem. Soc. Rev. 2014, 43, 381−398. (7) Paton, K. R.; Varrla, E.; Backes, C.; Smith, R. J.; Khan, U.; O’Neill, A.; Boland, C.; Lotya, M.; Istrate, O. M.; King, P.; Higgins, T.; Barwich, S.; May, P.; Puczkarski, P.; Ahmed, I.; Moebius, M.; Pettersson, H.; Long, E.;
D
DOI: 10.1021/acs.analchem.5b00228 Anal. Chem. XXXX, XXX, XXX−XXX
Letter
Analytical Chemistry Coelho, J.; O’Brien, S. E.; McGuire, E. K.; Sanchez, B. M.; Duesberg, G. S.; McEvoy, N.; Pennycook, T. J.; Downing, C.; Crossley, A.; Nicolosi, V.; Coleman, J. N. Nat. Mater. 2014, 13, 624−630. (8) Lee, S.; Cho, D.; Drzal, L. T. J. Mater. Sci. 2005, 40, 231−234. (9) Jeon, I. Y.; Shin, Y. R.; Sohn, G. J.; Choi, H. J.; Bae, S. Y.; Mahmood, J.; Jung, S. M.; Seo, J. M.; Kim, M. J.; Wook Chang, D.; Dai, L.; Baek, J. B. Proc. Natl. Acad. Sci. 2012, 109, 5588−5593. (10) Xie, J.; Spallas, J. P. Different contrast mechanisms in SEM imaging of graphene, Technical Report, Agilent Technologies; http://www.toyo.co. jp/file/pdf/spm/files/agi/5991-0782EN.pdf, 2012. (11) Wang, G.; Yang, J.; Park, J.; Gou, X.; Wang, B.; Liu, H.; Yao, J. J. Phys. Chem. C 2008, 112, 8192−8195. (12) Ferrari, A. C.; Meyer, J. C.; Scardaci, V.; Casiraghi, C.; Lazzeri, M.; Mauri, F.; Piscanec, S.; Jiang, D.; Novoselov, K. S.; Roth, S.; Geim, A. K. Phys. Rev. Lett. 2006, 97, 187401. (13) Chae, S. J.; Gunes, F.; Kim, K. K.; Kim, E. S.; Han, G. H.; Kim, S. M.; Shin, H. J.; Yoon, S. M.; Choi, J. Y.; Park, M. H.; Yang, C. W.; Pribat, D.; Lee, Y. H. Adv. Mater. 2009, 21, 2328−2333. (14) Reina, A.; Jia, X. T.; Ho, J.; Nezich, D.; Son, H. B.; Bulovic, V.; Dresselhaus, M. S.; Kong, J. Nano Lett. 2009, 9, 30−35. (15) Barron, A. R.; Flood, D. J. XPS of Carbon Nanomaterials. In OpenStax CNX, 2013; http://cnx.org/contents/[email protected]:10/Carbon_Nanotubes. (16) Gabbott, P. Principles and Applications of Thermal Analysis; Wiley: New York, 2008. (17) Haines, P. J. Principles of Thermal Analysis and Calorimetry; Royal Society of Chemistry: Cambridge, U.K., 2002. (18) Loryuenyong, V.; Totepvimarn, K.; Eimburanapravat, P.; Boonchompoo, W.; Buasri, A. Adv. Mater. Sci. Eng. 2013, 2013, 5. (19) Kishore, M.; Srimathi, P.; Kumar, S.; Addepalli, S.; Swaminathan, S.; Tilak, V.; Colborn, R. Bull. Mater. Sci. 2013, 36, 667−672. (20) Chen, W.; Yan, L. Nanoscale 2010, 2, 559−563. (21) Chabot, V.; Kim, B.; Sloper, B.; Tzoganakis, C.; Yu, A. Sci. Rep. 2013, 3, 1378. (22) Parviz, D.; Das, S.; Ahmed, H. S. T.; Irin, F.; Bhattacharia, S.; Green, M. J. ACS Nano 2012, 6, 8857−8867. (23) Buzaglo, M.; Shtein, M.; Kober, S.; Lovrincic, R.; Vilan, A.; Regev, O. Phys. Chem. Chem. Phys. 2013, 15, 4428−4435. (24) Gallego, J.; Batiot-Dupeyat, C.; Mondragón, F. J. Therm. Anal. Calorim. 2013, 114, 597−602. (25) Cataldo, F. Fullerenes, Nanotubes, Carbon Nanostruct. 2002, 10, 293−311. (26) Sahle-Demessie, E.; Zhao, A.; Salamon, A. W. A Study of Aged Carbon Nanotubes by Thermogravimetric Analysis; PerkinElmer, Inc.: Shelton, CT, 2011. (27) Mahajan, A.; Kingon, A.; Kukovecz, Á .; Konya, Z.; Vilarinho, P. M. Mater. Lett. 2013, 90, 165−168. (28) León, V.; Rodriguez, A. M.; Prieto, P.; Prato, M.; Vázquez, E. ACS Nano 2013, 8, 563−571. (29) Schawe, J. E. K. Elastomers; Mettler-Toledo: Columbus, OH, 2002; Vol. 2. (30) Memon, N. K.; Tse, S. D.; Al-Sharab, J. F.; Yamaguchi, H.; Goncalves, A. M. B.; Kear, B. H.; Jaluria, Y.; Andrei, E. Y.; Chhowalla, M. Carbon 2011, 49, 5064−5070. (31) Loomis, J.; Fan, X.; Khosravi, F.; Xu, P.; Fletcher, M.; Cohn, R. W.; Panchapakesan, B. Sci. Rep. 2013, 3, 1900. (32) Duhee Yoon, H. M. A. H. C. J. Korean Phys. Soc. 2009, 55, 1299− 1303. (33) Cançado, L. G.; Jorio, A.; Ferreira, E. H. M.; Stavale, F.; Achete, C. A.; Capaz, R. B.; Moutinho, M. V. O.; Lombardo, A.; Kulmala, T. S.; Ferrari, A. C. Nano Lett. 2011, 11, 3190−3196. (34) Nan, H. Y.; Ni, Z. H.; Wang, J.; Zafar, Z.; Shi, Z. X.; Wang, Y. Y. J. Raman Spectrosc. 2013, 44, 1018−1021. (35) Green, A. A.; Hersam, M. C. J. Phys. Chem. Lett. 2009, 1, 544−549. (36) Seo, J. W. T.; Green, A. A.; Antaris, A. L.; Hersam, M. C. J. Phys. Chem. Lett. 2011, 2, 1004−1008.
E
DOI: 10.1021/acs.analchem.5b00228 Anal. Chem. XXXX, XXX, XXX−XXX
