PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of scirepAboutEditorial BoardFor AuthorsScientific Reports
 
Sci Rep. 2017; 7: 45165.
Published online 2017 March 27. doi:  10.1038/srep45165
PMCID: PMC5366877

Degree of functionalisation dependence of individual Raman intensities in covalent graphene derivatives

Abstract

Covalent functionalisation of graphene is a continuously progressing field of research. The optical properties of such derivatives attract particular attention. In virtually all optical responses, however, an enhancement in peak intensity with increase of sp3 carbon content, and a vanishing of the peak position shift in monolayer compared to few-layer systems, is observed. The understanding of these seemingly connected phenomena is lacking. Here we demonstrate, using Raman spectroscopy and in situ electrostatic doping techniques, that the intensity is directly modulated by an additional contribution from photoluminescent π-conjugated domains surrounded by sp3 carbon regions in graphene monolayers. The findings are further underpinned by a model which correlates the individual Raman mode intensities to the degree of functionalisation. We also show that the position shift in the spectra of solvent-based and powdered functionalised graphene derivatives originates predominantly from the presence of edge-to-edge and edge-to-basal plane interactions and is by large functionalisation independent.

The optical properties of chemically functionalised graphene derivatives1,2,3,4,5,6,7,8,9,10 provide a platform with a broad range of possibilities for producing novel technologically relevant 2D carbon-based materials11,12,13. The absorption1,2, transmission3, fluorescence4,5, photoluminescence1,2,6,7 and Raman8,9,10 spectra of covalently functionalised graphene have been found to depend on the degree of functionalisation θ, defined as the ratio of C(sp3) to C(sp2) content in the material14. In particular, a significant intensity enhancement in the response signal with θ has been observed in various functionalised graphene derivatives1,2,7,8,15. Furthermore, a characteristic shift of the peak positions with θ was reported in solvent-based1,16,17 and bulk (powdered) functionalised graphene18, while the positions of the individual peaks remained constant for restacked monolayer derivatives1. The origin of both of the phenomena, and whether or not the two are connected, is under general debate. Moreover, the aforementioned spectroscopic methods are powerful tools not only for analysing the optical properties of carbon-based materials19; they also provide the operational range for graphene in potential optoelectronic devices20.

The introduction of chemical moieties in graphene leads to an sp2-to-sp3 rehybridization of lattice carbon atoms and the formation of π-conjugated domains of varying shapes and sizes6,21. Within the localized sp2 domains local bandgaps are predicted21 with energies depending on the domain area21. Such domains are generally considered to be carbon-based photoluminescent centres22,23 and can be expected to have a strong impact on the interaction between light and graphene derivatives.

We selected Raman spectroscopy, including a combined in situ Raman-spectroscopy and dry electrostatic doping method, as a principally non-invasive method for probing the (localized) sp2 areas of differently functionalised graphene derivatives24,25,26,27. Here we show that for monolayer functionalised graphene, the increase in optical response with θ is directly associated with the presence of photoluminescent C(sp2) domains in the material. We also propose a model which allows to quantify the degree of functionalisation by using the individual Raman mode intensities. Furthermore, we demonstrate that while the coupling of the photoluminescence affects the intensity of the optical response in few-layered restacked homogeneous systems, the shift in peak position can be attributed to edge-to-edge and edge-to-basal plane interactions in (disordered) solvent-based systems and powders.

Results

Individual Raman intensities in monolayer graphene derivatives

We refer to covalently modified monolayer graphene obtained by different reaction routes (cf. Methods) to cover a wide range of θ. Figure 1a depicts the schematic representation of these derivatives: pristine CVD graphene (CVD-G), hexylated (hexyl-G)24, mixed functionalised (hexyl/aryl-G)24, reduced oxo-graphene (r-oxo-G)25,26, arylated (aryl-G)27 as well as oxo-graphene (oxo-G)25,26 (Supplementary note 1). In order to eliminate the impact of the addends on the optoelectronic response, the chemical moieties used were either non-Raman active28 or showed a Raman response far from the dominant graphene peak positions29.

Figure 1
Photoluminescence modulated Raman intensity for different monolayer graphene derivatives.

The Raman spectra were taken at fixed experimental conditions (cf. Methods) and in the central areas of the flakes. For the purpose of this study, we focused on the dependence on the degree of functionalisation of the individual Raman D- and G-mode intensities, ID and IG, respectively (defined as the integrated peak areas, cf. Supplementary note 2). While the positions of both peaks remained constant (cf. Supplementary Fig. 1), a continuous increase in both ID and IG with the basal C(sp3) centre content was observed (Fig. 1b and c), spanning several orders of magnitude. Although our samples were equipped with different moieties, remarkably no significant variation in neighbouring data points was found. This indicates that the specific Raman response is functionality-independent and driven by the sp2 domains. The PL intensity originating from these domains, IPL, can be described as22

An external file that holds a picture, illustration, etc.
Object name is srep45165-m1.jpg

where E is the average band-gap of the sp2 domains (directly dependent on θ23), m an exponent associated with the density of states of the π-conjugated domains and fB(E) the Boltzmann distribution function. Following our initial assumption that the modulation of the optical response is a general effect, a specific coupling of the Raman signal to the PL can be expected, and thus the D-mode intensity to the lowest order can be written as

An external file that holds a picture, illustration, etc.
Object name is srep45165-m2.jpg

and the G-mode intensity

An external file that holds a picture, illustration, etc.
Object name is srep45165-m3.jpg

with I0 corresponding to the G-peak present in non-functionalised graphene. D0 and G0 account for the coupling strength of the D and G mode to the PL, respectively. It should be noted that all the parameters (including kD, kG, γD, γG) in equations (2) and (3) are dependent on the laser power, exposure time and temperature23. That is, a given set of parameter values is exclusively valid for a specific experimental condition during the Raman measurement. Therefore, the application of our model to a different experimental configuration requires a straightforward adjustment of the parameters in Eqs (2) and (3) (cf. Supplementary note 3).

The experimental data in Fig. 1b and c can be well reproduced within our model even considering the lower and upper uncertainty limits (cf. Table 1). This demonstrates that the experimentally observed Raman intensity increase can indeed be described as being modulated through PL-active sp2 domains.

Table 1
Values of the fitting parameters for D- and G-mode intensity obtained for the data in Fig. 3b and c.

Raman intensity and in situ electrostatic doping of oxo-G

To provide further evidence of the coupling between PL and the Raman response, we studied the dependence of ID and IG on the π-state occupation for the case of oxo-G (θ ~ 50%) at room temperature. A single oxo-G flake25 was lithographically contacted with Ti/Au electrodes on a SiO2 (300 nm) substrate (cf. Supplementary note 4). The Si backside of the chip was used as gate electrode. By applying a gate voltage, VG, the charge density in the flake was varied by up to ±1.4 × 1013 cm-2 (cf. Supplementary note 5). This corresponds to a shift of chemical potential, Δμ, by about ±370 meV, respectively. Raman spectra were taken at different gate voltages (schematic setup shown in Fig. 2a). This way, any optical transitions were directly tuned via electrostatic doping30 while preserving the structural integrity of the graphene.

Figure 2
Change in Raman intensity with electrostatic doping of monolayer oxo-G (θ = 50%).

In Fig. 2b, the Raman spectrum of oxo-G at zero gate voltage is shown. The evolution of the D- and G-mode intensities with varying chemical potential is shown in Fig. 2c. Extraction of electrons from oxo-G leads to a continuous increase of the D-mode intensity, which amounts to the total change of about 65% over the entire range of Δμ. In contrast, the G-mode intensity is hardly varying with the chemical potential. The latter implies that no chemical reduction occurred in the material during the experiment since the G-mode intensity is proportional to the number of lattice C(sp2) atoms illuminated by the laser beam30. Noteworthy, considering an average π-conjugated domain size in oxo-G of approximately 1 nm2 25,31, we can estimate that the addition of an extra electron will require 1 to 2 eV. This energy is of the same order of magnitude as the laser energy used (2.3 eV) implying that PL excitations are energetically possible. Also, this energy represents basically an activation energy which is well above the thermal energy at ambient conditions and thus consistent with the symmetric and nonlinear current-voltage characteristics32 (inset Fig. 2b). All these findings provide the final proof that the Raman intensity modulation originates from the photoluminescent C(sp2) domains, since electrostatic doping only changes the π-state occupancy.

Comparison of the intensity-based model to existing approaches

The consistency between our experimental data and our model shows that the functionalisation degree θ of a monolayer graphene can be unambiguously determined by solely relying on the actual intensities, ID and IG. Our model is also commensurate with the existing approaches which refer to the ratio of the Raman D- and G-modes, ID/IG, as a typical measure for the number of defects and disorder introduced14,33,34. This is demonstrated in Fig. 3a where the fitting of the ID/IG ratios with equations (2) and (3) recovers the mutual θ-dependence as reported in literature14,33,34. We note that for larger θ the uncertainty in the fitted ID/IG ratios tends to increase, which suggests that in this regime higher-order terms in equations (2) and (3) might have to be taken into consideration.

Figure 3
Correspondence between the intensity-based model with the ID/IG approach for determining degree of functionalisation θ.

We remark that the ratio ID/IG combined with the geometrical models is well known to be a limited description: First, no material-specific parameters are accounted for. Second, considering the addend related formation of π-conjugated domains23, this method only allows to deduce an average defect distance LD14,33,34 which is exclusively a good descriptive quantity for systems with a low amount of defects (well isolated basal C(sp3) centres). In other words, the ID/IG ratio is only suitable in the limit of a low θ (below ca. 1%). Third, no unique solution for LD is provided. And fourth, ID/IG leads to a loss of information which is contained in the individual ID and IG values and thus gives an incomplete measure regarding θ. This is illustrated in Fig. 3b,c for hexyl-G during laser induced defunctionalisation: With laser exposure time, both ID and IG values decreased by 35 to 40%. In contrast, the ID/IG ratio was reduced by less than 8% (inset Fig. 3c). The use of our model therefore is beneficial over the current ID/IG approaches.

Raman spectra in multilayered systems

Having established a direct link between PL-active sp2 domains and Raman intensity for individual monolayers of functionalised graphene, it is instructive to question how our findings translate to solution-processed bulk materials. In particular, we address the occurrence of a position shift in the optoelectronic response observed in powdered and solvent-based graphene derivative systems1,16,17,18. For this, we first chose regularly stacked oxo-G layers (θ ~ 50%) transferred on top of highly oriented pyrolytic graphite (HOPG) (Fig. 4a, cf. Methods). The Raman response was measured in the central areas (off-edge) of the flakes for different numbers of layers which were identified by AFM profile imaging (Fig. 4b,c). That is, here the areas close to the edges of the flakes were deliberately excluded from the analysis. With the used 532 nm laser excitation at 1 mW power, a penetration depth of at least 30 layers into the HOPG is expected based on the transmittance of single layer graphene35 (Supplementary Note 6). Representative Raman spectra for HOPG, mono-, bi-, tri- and ≥four-layer oxo-G are given in Fig. 4d showing an increase in signal intensities with each added layer, however, no shift in peak position was observed.

Figure 4
Evolution of the off-edge Raman response of stacked oxo-G layers with the layer number.

Accounting for the polydispersity of the system, we employed SRS36 to increase the statistical significance of the obtained data. From these measurements the average ID and IG values were extracted as function of layer number (Fig. 4e). For both, the peak intensities increase linearly, which shows that there are no significant interactions between adjacent oxo-G layers which might lead to the blocking of optical paths. However, the increments differ for ID and IG, which indicates that the PL contributions in a stacked system are not purely additive. Therefore, for the very specific conditions of well-aligned, non-interacting systems with uniform functionalisation a simple extrapolation from single layer analysis to multi-layers using Raman spectroscopy should be feasible.

Importantly, the fixed position of the Raman peaks suggests the shift observed in non-aligned systems1,16,17,18 is not associated with the photoluminescent domains and is of another origin. To corroborate this, we studied the positions of the D- and G-modes in few-layer oxo-G and r-oxo-G films restacked on a SiO2 surface (cf. Methods). Figure 5a and b show Raman mode position maps of the D- and G-modes, respectively, for a oxo-G film (θ = 50%). Analogous data for r-oxo-G (θ = 0.07%) are shown in Fig. 5c and d. For oxo-G films no clear distinction of the mode positions between centre and edge regions could be identified due to the strong broadening of the individual Raman peaks. In contrast, while the ID and IG positions in the central area of the r-oxo-G flakes remain constant irrespective of the layer number, a distinct shift can be observed at and close to the flake edges. Considering that the oxo-G at θ = 50% has on average significantly smaller π-conjugated domains with associated stronger PL contribution at our laser lines than r-oxo-G at θ = 0.07%, the reported shift in position of peaks in the optical response of solvent-based and powdered functionalised graphene is primarily not associated with the sp2 domains. Therefore, it can be attributed to edge-to-edge and edge-to-layer (basal plane) interactions within such polydisperse systems.

Figure 5
Dominant Raman mode positions in restacked functionalised layers on SiO2 substrate.

Discussion

Our combined study using Raman and electrostatic doping provides evidence that the generally observed modulation of the optical response in graphene derivatives is associated with photoluminescent sp2 domains, which form during chemical modification. Based on our findings, we demonstrate that the degree of functionalisation in covalently modified monolayer graphene can be directly related to the absolute Raman D- and G-mode intensities, and propose a phenomenological model which eliminates the ambiguity inherent to the geometrical ID/IG ratio approaches, therefore making analysis of graphene derivatives less ambiguous. Furthermore, we show that a similar intensity-based approach should be feasible to quantify the defect density in the case of few-layered, well-ordered systems. However, the presence of edges, interlayer interactions and different stacking geometries in a (disordered) bulk or solvent-based system can lead to a position shift in the optical response. Our results also provide insight into the optical response of graphene quantum dots and assemblies of those where edge-to-edge interactions can play a role. Finally, the coupling of PL to Raman excitations can be expected to occur in other covalently functionalised layered 2D materials, too, and is probably even enhanced if they are inherently PL active unlike graphene.

Methods

Covalent functionalisation of graphene

Oxo-graphene (oxo-G) of 50% was synthesized from natural graphite in sulfuric acid with potassium permanganate as oxidant25; for 6% oxo-G the starting material was graphite sulphate26. Arylated CVD graphene (aryl-G): Bis-(4-tert-butylphenyl) iodonium hexafluorophosphate was deposited from solution (CH2Cl2) on a monolayer flake. The reaction of 4-tert-butylphenyl (tBP) cations was subsequently activated by a laser pulse (532 nm, 5 s, 10 mW) within the Raman spectrometer. Generated phenyl cations reacted with the carbon lattice of graphene by covalent bond formation27. The reaction scheme is shown in Fig. S1. Hexylated CVD graphene (hexyl-G), hexylated and arylated CVD graphene (hexyl/aryl-G): were synthesized according to Ref. 24. Laser-induced thermal defunctionalisation of hexyl-G was carried out with the laser line 532 nm and a power of 15 mW. The CVD graphene (cf. Supplementary Fig. 2 for Raman spectrum) was grown on Cu and purchased from Graphene Supermarket (https://graphene-supermarket.com/CVD-grown-graphene).

Raman spectroscopy

Raman spectroscopic characterization was carried out on a Horiba Jobin Yvon LabRAM Aramis confocal Raman microscope (excitation wavelength: 532 nm) with a laser spot size of ~1 μm (Olympus LMPlanFl 100x, NA 0.90). Raman spectra were taken with a 532 nm laser line, 1 s acquisition time, 1 mW power, at room temperature. Statistical Raman measurements were obtained through a motorized x-y table in a continuous linescan mode (SWIFT-module).

Preparation of Langmuir-Blodgett films

Langmuir-Blodgett Films of oxo-G on HOPG and SiO2 were prepared using Langmuir-Blodgett Minitrough from KSV NIMA. Oxo-G was dissolved in methanol/water mixtures. The subphase was water. Oxo-G transfer was carried out at the pressure of 10 mN m−1. Oxo-G was subsequently reduced by vapor of a 1:1 volume mixture of hydriodic acid and trifluoroacetic acid at 80 °C.

Data availability

All relevant data are available from the authors on request.

Additional Information

How to cite this article: Vecera, P. et al. Degree of functionalisation dependence of individual Raman intensities in covalent graphene derivatives. Sci. Rep. 7, 45165; doi: 10.1038/srep45165 (2017).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Material

Supplementary Information:

Acknowledgments

The authors acknowledge financial support from the Deutsche Forschungsgemeinschaft (DFG-SFB 953 “Synthetic Carbon Allotropes”, A1, B12), the European Research Council (ERC grant no. 259286), and the Graduate School Molecular Science (GSMS) for financial support. The research leading to these results has received partial funding from the European Union Seventh Framework Programme under grant agreement no. 604391 Graphene Flagship. S.E. acknowledges the DFG for financial support by grant no. EI 938/3-1.

Footnotes

The authors declare no competing financial interests.

Author Contributions S.E., F.H., V.K. and A.H. supervised the project. P.V. provided the Raman spectroscopy. R.S. carried out the functionalisation of CVD graphene. M.K.G. and V.K. developed the model. M.K.G. and V.K. developed the in situ Raman and electrostatic gating set-up and carried out the measurements. A.V. and J.M. contributed to the Raman measurements. All authors discussed the results. P.V., M.K.G., S.E., F.H., and V.K. took lead in manuscript write-up.

References

  • Mei Q. et al. . Highly efficient photoluminescent graphene oxide with tunable surface properties. Chem. Commun. 46, 7319–7321 (2010). [PubMed]
  • Eda G. et al. . M. Blue Photoluminescence from Chemically Derived Graphene Oxide. Adv. Mater. 22, 505–509 (2010). [PubMed]
  • Bonaccorso F., Sun Z., Hasan T. & Ferrari A. C. Graphene photonics and optoelectronics. Nature Photon. 4, 611–622 (2010).
  • Schäfer R. A. et al. . On the way to graphane: pronounced fluorescence of polyhydrogenated graphene. Angew. Chem. Int. Ed. 52, 754–757 (2013). [PubMed]
  • Naumov A. et al. . Graphene oxide: a one- versus two-component material. J. Am. Chem. Soc. 138, 11445–11448 (2016). [PubMed]
  • Strauss V., Schäfer R. A., Hauke F., Hirsch A. & Guldi D. M. Polyhydrogenated graphene: excited state dynamics in photo- and electroactive two-dimensional domains. J. Am. Chem. Soc. 137, 13079–13086 (2015). [PubMed]
  • Luo Z., Vora P. M., Mele E. J., Johnson A. T. C. & Kikkawa J. M. Photoluminescence and band gap modulation in graphene oxide. Appl. Phys. Lett. 94, 111909 (2009).
  • Jang A. et al. . Reversibly Light-Modulated Dirac Point of Graphene Functionalized with Spiropyran. ACS Nano, 6, 92017–9213 (2012). [PubMed]
  • Zhong X. et al. . Aryne cycloaddition: highly efficient chemical modification of graphene. Chem. Commun. 46, 7340–7342 (2010). [PubMed]
  • Frolova L. V. et al. . Tetracyanoethylene oxide-functionalized graphene and graphite characterized by Raman and Auger spectroscopy. Carbon 81, 216–222 (2015). [PMC free article] [PubMed]
  • Hirsch A., Englert J. M. & Hauke F. Wet chemical functionalization of graphene. Acc. Chem. Res. 46, 87–96 (2013). [PubMed]
  • Johns J. E. & Hersam M. C. Atomic covalent functionalization of graphene. Acc. Chem. Res. 46, 77–86 (2013). [PMC free article] [PubMed]
  • Sarkar S., Bekyarova E. & Haddon R. C. Chemistry at the Dirac Point: Diels-Alder Reactivity of Graphene. Acc. Chem. Res. 45, 673–682 (2012). [PubMed]
  • Englert J. M. et al. . Scanning-Raman-microscopy for the statistical analysis of covalently functionalized graphene. ACS Nano 7, 5472–5482 (2013). [PubMed]
  • Sun X. et al. . Nano-Graphene Oxide for Cellular Imaging and Drug Delivery, Nano Res. 1, 203–212 (2008). [PMC free article] [PubMed]
  • Chien C.-T. et al. . Tunable Photoluminescence from Graphene Oxide. Angew. Chem. Int. Ed. 51, 6662–6666 (2012). [PubMed]
  • Geng J. & Jung H.-T. Porphyrin Functionalized Graphene Sheets in Aqueous Suspensions: From the Preparation of Graphene Sheets to Highly Conductive Graphene Films. J. Phys. Chem. C 114, 8227–8234 (2010).
  • Eda G., Fanchini G. & Chhowalla M. Large-area ultrathin films of reduced graphene oxide as a transparent and flexible electronic material. Nature Nanotech. 3, 270–274 (2008). [PubMed]
  • Ferrari A. C. & Basko D. M. Raman spectroscopy as a versatile tool for studying the properties of graphene. Nat. Nanotechnol. 8, 235–246 (2013). [PubMed]
  • Zhang H. et al. . Aryl functionalization as a route to band gap engineering in single layer graphene devices. Nano Lett. 11, 4047–4051 (2011). [PubMed]
  • Robertson J. & O’Reilly E. P. Electronic and atomic structure of amorphous carbon. Phys. Rev. B 35, 2946–2957 (1987). [PubMed]
  • Demichelis F., Schreiter S. & Tagliaferro A. Photoluminescence in a-C:H films. Phys. Rev. B 51, 2143–2147 (1995). [PubMed]
  • Robertson J. Recombination and photoluminescence mechanism in hydrogenated amorphous carbon. Phys. Rev. B 53, 16302–16305 (1996), [PubMed]
  • Knirsch K. C., Schäfer R. A., Hauke F. & Hirsch A. Mono- and Ditopic Bisfunctionalization of Graphene. Angew. Chem. Int. Ed. 55, 5861–5864 (2016). [PubMed]
  • Eigler S. et al. . Wet chemical synthesis of graphene. Adv. Mater. 25, 3583–3587 (2013). [PubMed]
  • Eigler S. Graphite sulphate – a precursor to graphene. Chem. Commun. 51, 3162–3165 (2015). [PubMed]
  • Merritt E. A. & Olofsson B. Diaryliodonium salts: a journey from obscurity to fame. Angew. Chem. Int. Ed. 48, 9052–9070 (2009). [PubMed]
  • Zucaro L., Kern K. & Balasubramanian K. Identifying Chemical Functionalization on Individual Carbon Nanotubes and Graphene by Local Vibrational Fingerprinting, ACS Nano 9, 3314–3323 (2015). [PubMed]
  • Niyogi S. et al. . Spectroscopy of Covalently Functionalized Graphene, Nano Lett. 10, 4061–4066 (2010). [PubMed]
  • Froehlicher G. & Berciaud S. Raman spectroscopy of electrochemically gated graphene transistors: Geometrical capacitance, electron-phonon, electron-electron, and electron-defect scattering. Phys. Rev. B 91, 205413 (2015).
  • Erickson K. et al. . Determination of the local chemical structure of graphene oxide and reduced graphene oxide. Adv. Mater. 22, 4467–4472 (2010). [PubMed]
  • Shklovskii B. I. & Efros A. L. Electronic Properties of Doped Semiconductors, (Springer-Verlag, Berlin Heidelberg, 1984).
  • Cançado L. G. et al. . Quantifying Defects in Graphene via Raman Spectroscopy at Different Excitation Energies. Nano Lett. 11, 3190–3196 (2011). [PubMed]
  • Lucchese M. M. et al. . Quantifying ion-induced defects and Raman relaxation length in graphene. Carbon 48, 1592–1597 (2010).
  • Nair R. R. et al. . Fine structure constant defines visual transparency of graphene. Science, 320, 1308–1308 (2008). [PubMed]
  • Wang Q. H. et al. . Understanding and controlling the substrate effect on graphene electron-transfer chemistry via reactivity imprint lithography. Nature Chem. 4, 724–732 (2012). [PubMed]

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group