Raman spectroscopy of graphene and graphite: Disorder, electron–phonon coupling, doping and nonadiabatic effects
Introduction
Graphene is the two-dimensional (2D) building block for carbon allotropes of every other dimensionality. It can be stacked into 3D graphite, rolled into 1D nanotubes, or wrapped into 0D fullerenes. Its recent discovery [1], [2], [3] completes the carbon-family. This finally opens the opportunity to study experimentally its electronic and phonon properties, which so far had to be inferred from theory.
In general, carbon-based materials play a major role in today’s science and technology and the discovery of graphene is the last of a long string of continuous advances in the science of carbon. These include, for example, the chemical vapour deposition of diamond [4], the discovery of fullerenes [5] and carbon nanotubes [6], [7], [8], and mastering the properties of amorphous and disordered carbons [9], [10], [11], [12], to span, on demand, almost all the range from graphite to diamond to carbon polymers [9], [10], [11], [12], [13], [14], [15], [16], [17]. Indeed, amorphous and diamond-like carbons (DLC) are currently used in many every-day life applications, such as, for example, magnetic hard disk coatings, wear protective and anti-reflective coatings for tribological tools, engine parts, razor blades and sunglasses, biomedical coatings (such as hips or stents) and microelectro-mechanical systems [9], [10]. Graphitic carbon and, to an extent, carbon nanotubes, are also utilized in batteries [18]. Applications in field emission displays, microwave amplifiers, transistors, supercapacitors, structural and conductive composites, photonic devices although all well beyond the proof-of-principle stage, have yet to make it to market. One of the main advantages of graphene is the possible advent of a planar technology, compatible with existing manufacturing processes [1].
A key requirement for carbon research and applications is the ability to identify and characterize all the members of the carbon family, both at the lab- and at mass-production scale. To be appealing, a characterization tool must be nondestructive, fast, with high resolution and give the maximum structural and electronic information. Raman spectroscopy provides all these. It is the backbone of research in such diverse fields, ranging from physics, to engineering, chemistry and biology. Indeed, most of the papers published every year on carbon materials have at least a Raman spectrum in them [19].
Raman spectroscopy can thus become the standard also in the fast growing field of graphene. One should remember that in the process of making graphene, be it from mechanical cleavage [1], [2], “expitaxial growth” [20], [21], chemical vapour deposition [21], [22], chemical exfoliation [23], all sorts of carbon species can in principle occur, similarly to what happens when making nanotubes. Unwanted by-products and structural damage can also be created while shaping graphene into devices. It would thus be advisable to have a structural reference, monitored, for example, by Raman spectroscopy, as common denominator to compare the materials used by different groups. This is standard practice in the field of nanotubes and amorphous and diamond-like carbons [19]. In the case of nanotubes, other optical structural characterization techniques, such as photoluminescence excitation spectroscopy, are now very popular [24], as ellipsometry or XPS are in amorphous and diamond like carbons [25]. It is thus foreseeable that other optical techniques (maybe even simpler than Raman spectroscopy) will become available also for graphene, once the field expands experimentally.
The toll for the simplicity of Raman measurements is paid when it comes to spectral interpretation. The Raman spectra of all carbon systems show only a few prominent features, no matter the final structure, be it a conjugated polymer or a fullerene [19]. The spectra appear deceivingly simple: just a couple of very intense bands in the 1000–2000 cm−1 region and few other second-order modulations. However, their shape, intensity and positions allow to distinguish a hard amorphous carbon, from a metallic nanotube, giving as much information as that obtained by a combination of other lengthy and destructive approaches [19]. The peculiar dispersion of the electrons in graphene is the fundamental reason why Raman spectroscopy in carbons is always resonant and, thus, a powerful and efficient probe of their electronic properties, not only of their vibrations [19]. This explains why the interpretation of the Raman spectra of graphitic materials was investigated for almost 40 years [19], [26] and why intense effort has been put towards the Raman measurement of few-layer graphite samples [27], [28], even before the discovery of graphene.
Section snippets
and peaks, double resonance and Kohn anomalies
Fig. 1 compares the Raman spectra of a few representative carbon materials: graphite, metallic and semiconducting nanotubes and high and low sp3 amorphous carbons, all measured for visible excitation. Fig. 2 plots the peak position as a function of excitation energy for defected graphite.
The main features in the Raman spectra of carbons are the so-called and peaks, which lie at around 1560 and 1360 cm−1 respectively for visible excitation. In amorphous carbons a peak at around 1060 cm−1 (
Electron–phonon coupling from phonon dispersions and Raman line widths
Electron–phonon coupling (EPC) is a key physical parameter in graphene and nanotubes. Ballistic transport, superconductivity, excited state dynamics, Raman spectra and phonon dispersions all fundamentally depend on it. In nanotubes, the optical phonons EPC are also extremely relevant since electron scattering by optical phonons sets the ultimate limit to high field ballistic transport [56], [57], [58], [59], [60]. Many tight-binding calculations of optical phonons EPC in graphene and nanotubes
The Raman spectrum of graphene and graphene layers
Fig. 5 compares the Raman spectra of graphene and bulk graphite measured at 514.5 nm excitation [29]. The two most intense features are the peak at 1580 cm−1 and a band at ∼2700 cm−1, historically named , since it is the second most prominent band always observed in graphite samples [38]. However we now know that this band is the second order of the peak. Thus we believe it is more convenient to refer to it as peak [29]. Fig. 5 also shows another peak at ∼3250 cm−1. Its frequency is
The Raman spectrum of doped graphene: Breakdown of the adiabatic Born–Oppenheimer approximation
Doping changes the Fermi surface of graphene. This moves the Kohn anomaly away from . Thus, since first order non-double resonant Raman scattering probes phonons, intuitively we expect a stiffening of the peak. Indeed, refs. [30], [79] reported that the peak of graphene responds to doping. The doping level was controlled by applying a gate voltage. The peak upshifts for both holes and electron doping [30], [79].
Fig. 8(a) and (b) reports the peak position and FWHM measured at
Disordered graphite and graphene
In order to compare different samples and devices, or different locations on the same sample, another crucial parameter, besides doping, is the amount of disorder. For multilayers assessing turbostraticity is also important.
We introduced a three-stage classification of disorder, leading from graphite to amorphous carbons [13], [14], which allows to simply assess all the Raman spectra of carbons. The Raman spectrum is considered to depend on:
- (i)
clustering of the sp2 phase;
- (ii)
bond disorder;
- (iii)
presence of
Edges and ribbons
The sample edges can be always seen as defects. Thus, when the laser spot includes them, even if the bulk sample is perfect, a peak will appear. Fig. 5 shows no peak at the centre of a typical graphene layer, proving the absence of a significant number of defects in the structure. A single peak is only observed at the sample edge, Fig. 11, consistent with the single peak discussed in Section 4. On the other hand the peak at the edge of graphite consists of two peaks and [36],
Conclusions
A review of the Raman spectra of graphite and graphene was presented. The and Raman peaks change in shape, position and relative intensity with number of graphene layers. This reflects the evolution of the electronic structure and electron–phonon interactions. Doping upshifts and sharpens the peak for both both holes and electrons. Disorder can be monitored via the peak. Thus Raman spectroscopy can be efficiently used to monitor a number of layers, quality of layers, doping level and
Acknowledgements
The author acknowledges S. Pisana, M. Lazzeri, C. Casiraghi, V. Scardaci, S. Piscanec, K.S. Novoselov, A.K. Geim, F. Mauri, J.C. Meyer, J. Robertson and funding form the Royal Society and The Leverhulme Trust.
References (101)
- et al.
J. Cryst. Growth
(1976) - et al.
Carbon
(2006) Trib. Int.
(2004)Surf. Coat. Technol.
(2004)- et al.
Diam. Relat. Mater.
(2001) - et al.
J. Non-Cryst. Solids
(1980) - et al.
Solid State Commun.
(1981) - et al.
J. Non-Cryst. Solids
(1998) - et al.
Carbon
(1984) - et al.
Physica B
(1996)
Solid. State Commun.
Chem. Phys. Lett.
Solid State Commun.
Carbon
Chem. Phys. Lett.
Science
Proc. Natl. Acad. Sci.
Nat. Mater.
Science
Nature
Nature
Mater. Today
MRS Bull.
Phys. Rev. B
Phys. Rev. B
Phys. Rev. B
Phys. Rev. B
J. Appl. Phys.
Jpn. J. Appl. Phys.
J. Phys. Chem. B
J. Phys. C
Phys. Rev. B
Nature
Science
J. Chem. Phys.
Science
Phys. Rev. Lett.
Phys. Rev. Lett.
Nat. Mater.
Phys. Rev. B
Diam. Relat. Mater.
J. Chem. Phys.
Philos. Trans. R. Soc. Lond. Ser. A
Phys. Rev. B
Phys. Rev. B
Phys. Rev. Lett.
Opt. Spektrosk.
Phys. Rev. B
Phys. Rev. B
Cited by (6527)
Defect-initiated formation mechanism of 3D carbon tracks on flexible transparent substrates by laser irradiation
2024, Optics and Laser TechnologyStructural transformation of carbon-encapsulated core-shell CoNi nanoparticles during magnetically induced CO<inf>2</inf> reduction into CO
2024, Applied Catalysis B: EnvironmentalTribological properties of graphene-reinforced polyurethane bearing material
2024, Tribology International