a compilation by J. M. Walter

## Introduction

Raman investigations of carbon material are used to determine the degree of crystallinity from poorly ordered carbon to graphite and diamond. This page focusses on raman measurements of graphite with examples from the Central Zone of the Damara Belt in Namibia. It comprises shortly the experimental background, the evaluation of the raman data and the calculation of the graphite crystallinities.

## Experimental background

Raman spectroscopy is based on the Raman Effect, which originates from the scattering of electromagnetic radiation by the electron sheaths of the investigated molecules. The Raman Effect is virtually independent of the wavelength of the radiation beam. Therefore, a monochromatic laser beam is used for Raman spectroscopy. Approximately 99.99 % of the laser beam radiates through the sample, while a small part is scattered elastically by the molecules (Rayleigh scattering). An even smaller part (about 10-6 %) is scattered inelastically, which is referred to as the Raman scattering. This Raman radiation is emitted either in longer (Stokes radiation) or in shorter wavelengths (Antistokes radiation) than the original wavelength of the laser beam. This difference in frequency is expressed by wave numbers relative to the original frequency of the laser beam. The original frequency is therefore calibrated to zero. This Raman shift is specifically characteristic for different molecules.

The Raman shift for carbonaceous material (CM) is divided into first and second order regions after Tuinstra & Koenig (1970). The measured first-order region is shown in Figure 1. The first-order region lies in the range of 1100 to 1800 wave numbers (Δ cm-1), and the main graphite band - the G band - is at ~1582 Δ cm-1. This band is inherent in graphite lattices. For more poorly crystalline graphite additional bands are recognisable at ~1150 Δ cm-1, at ~1355 Δ cm-1, at ~1500 Δ cm-1 and at ~1622 Δ cm-1. The band at ~1355 Δ cm-1 is referred to as the main defect band (D1-band) (Fig. 1). This band occurs when defects are present in the carbon aromatic structure (Beny-Bassez & Rouzaud, 1985). It is also sensitive to graphite intercalations (Dresselhaus & Dresselhaus, 1982). The 1150 Δ cm-1 band appears only in very poorly organised CM (Beyssac et al., 2002). Also the band at 1500 Δ cm-1 (D3 band) is attributed to defects outside the plane of aromatic layers like tetrahedral carbons (Beny-Bassez & Rouzaud, 1985). It occurs as a wide band in poorly crystallised CM (Beyssac et al., 2002). The D2 band at ~1622 Δ cm-1 appears as a shoulder peak of the G band and is also absent in highly crystalline graphite (Fig. 1).

Fig. 1 First order raman spectrum for graphite

The second order region from 2200 to 3400 Δ cm-1 includes several bands at ~2400 Δ cm-1, at ~2700 Δ cm-1, at ~2900 Δ cm-1 and at ~3300 Δ cm-1, depending on the degree of graphite crystallinity. The S1 band at ~2700 Δ cm-1 splits up into two bands at high crystallinities. It is therefore the most important indicator band for graphite crystallinities in the second order region.

In general, Raman spectra for well crystallised graphite include the existence of D1 and D2 bands in the first order region (Wopenka & Pasteris, 1993). To estimate the degree of graphite crystallinity, it is important to reveal some information about the variation of the G band position, about the full widths at half maximum of the G and D1 bands and about the intensity ratios between D1 and G. This intensity ratio can be used to quantify the crystallinity of graphite according to the linear relationship between the D1/G intensity ratio and 103/La. La quantifies after Tuinstra & Koenig (1970), the crystallinity of graphite. It quantifies the mean basal plane diameter of graphite parallel to (001) and is expressed in Å.

## Calculation of graphite crystallinity

Modern data aquisition software for raman spectroscopy also allows the necessary calculations for further data evaluation. After the background subtraction, a calculation of the full width at half maximum, the position and intensity of the maxima is necessary by fitting a calculated curve to the measured spectrum. In this project a Gauss-Lorentz fit obtained the best results for the graphite spectra (Fig. 1).

In general it should be kept in mind, that the raman effect varies in graphite with the crystallographic orientation of the measured graphite crystal.

A first approximation about the degree of graphite crystallinity is the G band position and accordingly also the variation of this G band position within a sample (Beny-Bassez & Rouzaud, 1985; Beyssac et al., 2002). For highly crystalline graphite the G band position should be at ~1582 Δ cm-1 with a variation of about 20 Δ cm-1. Both authors state, that a decrease in wave numbers for the G band position occurs for mean wave numbers from ~1590 Δ cm-1 to ~1582 Δ cm-1 with increasing crystallinities. The G band position of poorly-ordered graphite varies about 50 Δ cm-1 (Beyssac et al., 2002). An example of a plot of the G band position against sample numbers is shown in Figure 2 a.

Fig. 2 a to e show examples of plots characterising graphite crystallinities. They are plotted against sample numbers. a) G-band position, b) FWHM of G-band, c) FWHM of D1-band, d) D1/G intensity ratio, e) calculated graphite crystallinities.

A second indicative parameter for the degree of graphite crystallinity is the full width at half maximum (FWHM) of the G band. Similarly to the G band position a decrease in the FWHM from about 60 to 120 Δ cm-1 to about 22 Δ cm-1 is observable with increasing graphite crystallinities. Also the variability in the FWHM of the G band is representative for the crystallinity of graphite. For poorly ordered CM this variation may be up to 70 Δ cm-1, whereas highly ordered graphite shows a variation of about 5 to 10 Δ cm-1 (Beny-Bassez & Rouzaud, 1985; Wopenka & Pasteris, 1993; Beyssac et al., 2002). For an example see Figure 2 b.

Thirdly, the full width at half maximum of the D1 band gives a good indication for the degree of graphite crystallinities. After Beny-Bassez & Rouzaud (1985) a decrease from 260 Δ cm-1 to about 50 Δ cm-1 is observable with increasing graphite crystallinities from poorly to well ordered graphite. Accordingly, the variation of the FWHM for the D1 band has to be considered evaluating the graphite crystallinities. After Beny-Bassez & Rouzaud (1985) and Wopenka & Pasteris (1993) this variation is from about 120 Δ cm-1 for poorly ordered CM to about 25 Δ cm-1 for well ordered graphite. An example of such a plot is shown in Figure 2 c.

Graphite crystallinities are also characterised by the intensity ratio of the D1 band and the G band. This varies as much as one order of magnitude between the lowest and highest degrees of graphite crystallinity. After Wopenka & Pasteris (1993) and Beyssac et al. (2002) poorly-ordered carbonaceous material shows mean D1/G intensity ratios between 1 and 2.6 with a standard deviation of 0.8 to 1.2. For highly-crystalline graphite mean D1/G intensity ratios of 0.1 to 0.3 are measurable with a standard deviation of up to 0.2. The variation in the measurements of a sample for highly-crystalline graphite is about 0.4 (Wopenka & Pasteris, 1993; Beyssac et al., 2002). An example of the graphite D1/G intensity ratios are plotted against the sample numbers in Figure 2 d.

The D1/G intensity ratio is used to calculate the crystallinities of graphite. As described above, there is a linear relationship between this ratio and a factor of the mean basal plane diameter La (Tuinstra & Koenig, 1970; Beny-Bassez & Rouzaud, 1985; Wopenka & Pasteris, 1993). A plot of graphite crystallinities is shown in Figure 2 e. This relationship after Tuinstra & Koenig (1970) is shown in Figure 3.

Fig. 3 D1/G intensity ratio against graphite crystallinities after Tuinstra & Koenig (1970)

## For further information see also:

For further information see also:

Reich, S. & Thomsen, C. 2004. Raman spectroscopy of graphite. Philosophical Transactions: Mathematical, Physical & Engineering Sciences, November 15, 2004, vol. 362, no. 1824, pp. 2271-2288(18)

Tan, P., Dimovski, S., Gogotsi, Y. 2004. Raman scattering of non-planar graphite: arched edges, polyhedral crystals, whiskers and cones. Philosophical Transactions: Mathematical, Physical & Engineering Sciences, November 15, 2004, vol. 362, no. 1824, pp. 2289-2310(22)

References

Beny-Bassez, C. & Rouzaud, J. N. 1985. Characterization of carbonaceous materials by correlated electron and optical microscopy and Raman microspectroscopy. Scanning Electron Microscopy 1985(1), 119-132.

Beyssac, O., Rouzaud, J. N., Goffe, B., Brunet, F. & Chopin, C. 2002. Graphitization in a highpressure, low-temperature metamorphic gradient; a Raman microspectroscopy and HRTEM study. Contributions to Mineralogy and Petrology 143(1), 19-31.

Dresselhaus, M. S. & Dresselhaus, G. 1982. Light scattering in graphite intercalation compounds. In: Light scattering in solids (edited by Cardona, M. & Guntherodth, G.). Springer, New York, 3-57.

Tuinstra, F. & Koenig, J. L. 1970. Raman Spectrum of Graphite. Journal of Chemical Physics 53(3), 1126-1130.

Wopenka, B. & Pasteris, J. D. 1993. Structural characterization of kerogens to granulite-facies graphite; applicability of Raman microprobe spectroscopy. American Mineralogist 78(5-6), 533-557.