• Ei tuloksia

Raman spectroscopy is one of the main and efficient technique to determine the type, quality and property of a material, specially in case of carbon and its allotrope [66, 67]. The basic working principle of Raman spectroscopy is based on inelastic Raman scattering. Before we dwell into the details of Raman spectroscopy of graphene and graphitic carbon layers, let us brief out a little background on Raman spectroscopy in general and Raman scattering.

3.4.1 Raman Scattering

When incident light hits a medium, energy and direction of incident light may change. This is called scattering of light. Change in energy of light means change in frequency or wavelength of light. In the case of scattering of light, wavelength (or energy) change is typically given as reciprocal of wavelength or wavenumber in cm1 for convenience. If energy of light does not change, or the change is smaller than 105 cm1, scattering of light is called Rayleigh scattering. If energy of light changes more than 1 cm1, scattering of light is called Raman scattering [68]. The differ-ence between frequency of incident light and frequency of Raman scattered light is known as Raman shift or Raman frequency, which is typically presented using wavenumber. If wavenumber of incident light is set to zero, both positive and neg-ative Raman shifts are observed [68]. These are called anti-Stokes and Stokes lines respectively. Absolute value of Raman shift is the same for a Stokes and anti-Stokes line [69]. However, anti-Stokes lines have lower intensity. Furthermore, Raman shift does not depend on the frequency of incident light [68]. Energy difference between incident light and Raman scattered light corresponds to molecular vibration mode.

However, not all of the modes of a molecule cause Raman scattering. Mode is Ra-man active, i.e. it causes RaRa-man scattering, if it causes change in polarizability [69].

From quantum mechanical point of view molecule changes from one state to another through a virtual state in the case of Raman scattering. These virtual states do not exist in reality, they represent the case, where incident photon is annihilated and scattered photon is created at the same time. To distinguish Raman scattering from fluorescence, process of Raman scattering takes one picosecond or less, whereas flu-orescence involves absorbtion and emission and is considerably longer process [70].

The energy level diagram of the different scattering processes are shown in figure 3.4.

3.4.2 Raman Spectroscopy

Vibrational modes and Raman shifts are different for different molecules, therefore Raman scattering can be used in spectroscopy. Raman spectrometer consists of light source, optical microscope, spectral dispersion and spectral acquisition components.

Usually laser is used as a light source, because it is highly monochromatic, coherent and polarized and can be directed and focused easily. Monochromatic light source is important, because absolute value of Raman shift depends on the wavelength of in-cident light. If light source is not very monochromatic, peaks in Raman spectra may not be that clear. The optical microscope focusses incident light, so that it is more monochromatic, and also adjusts intensity and polarization of incident light. Spec-tral dispersion and acquisition parts form the Raman scattering spectrum known as Raman spectrum [68]. Raman spectrum is generally obtained in the form of in-tensity as a function of Raman shift. Commonly inin-tensity, frequency (Raman shift) and line shape and width are the spectral parameters seen in the spectrum. Ra-man spectra are different for different molecules, therefore sample’s molecules can be identified, provided Raman spectra of sample’s molecules are known. Polariza-tion of Raman scattered light depends on polarizaPolariza-tion of incident light and also on symmetry of sample’s molecules or crystal structure [68]. Raman spectroscopy is comparable to infrared spectroscopy. In infrared spectroscopy, intensities of inci-dent and transmitted light are compared to find the wavelengths absorbed by the sample’s molecules. Infrared spectroscopy also gives information about molecules’

vibrations, yet it has many differences with Raman spectroscopy. Vibrational modes can be either Raman active, infrared active or both. Also some modes can be seen better in Raman spectrum than in infrared spectrum and vice versa. Water and glass do not interfere measurements in Raman spectroscopy, whereas they absorb infrared radiation strongly, thus causing problems in infrared spectroscopy. Another advantage in Raman spectroscopy is that Raman spectrum from one measurement can cover large range of frequencies. For instance, gratings and detectors have to be changed in infrared spectroscopy to achieve such large range of frequencies. Dis-advantages in Raman spectroscopy are possible fluorescence and local heating, and equipment of Raman spectroscopy is generally more expensive [71]. The schematic of a Raman spectrometer is shown in figure 3.5.

Figure 3.4: Energy level diagram showing Rayleigh scattering, Raman scattering and infrared absorption.

Figure 3.5: Schematic of a Raman spectrometer.

3.4.3 Raman Spectroscopy of Graphite and Graphene

Analysis of Raman spectra determine the quality, nature and number of layers of the graphitic structure. Within the wavenumber range 1000 cm1 to 3500 cm1, graphene or in general graphitic carbon layers give three peaks namely, the 2D, G and D peaks. For any carbon material possessingsp2hybridization bonds, intense 2D and

G peaks are observed at ∼2700 cm1 and at ∼1580 cm1 respectively [66]. The D or the defective peak, usually obtained at∼1350 cm1 gives a qualitative measure of the disorder in the material structure [66]. The position, shape and relative intensity of the peaks vary, giving quite significant idea about the crystallinity and number of layers of the graphitic structure [72]. The ratio of the D and G peaks specifically indicates the crystallinity of the graphene/graphite sample, while on the other hand the position of the 2D peak gives idea about the thickness of the graphitic layers [73, 74]. In the Raman spectrum, a wide G peak between 1500 cm1 and 1550 cm1 signifies amorphous form of carbon. Also, shift of the G peak towards 1600 cm1 from its original position at 1580 cm1determines the presence of nanosized graphitic flakes while a shift towards 1500 cm1 determines the presence of amorphous carbon [75]. Similarly, the widening and shifting of the 2D peak illustrates the number of layers of the graphitic material. The 2D peak is located at 2730 cm1 in bulk graphite [76], while in graphene monolayer it is located at 2680 cm1 [77]. The 2D peak is typically located near 2700 cm1 for a few layered graphene [74, 77]. Also, it may also be noted here that apart from the aforesaid three prominent peaks, minor peaks may also be observed at locations near 1600 cm1, 3250 cm1 and 2450 cm1 that arises due to doping and defects in the material [66].