Raman spectroscopy probes the vibrations (phonons) in a material. These phonons are characteristic of a given material and allow its chemical and structural properties to be investigated. Raman spectroscopy can be employed in a wide range of application areas, including semiconductors, pharmaceuticals, and living cells. It can be used to investigate any substance (whether it be solid, liquid, or gas), except pure metals.
A technique has been developed for the rapid characterization of large regions of graphene—as big as cm2—enabling the number of graphene layers, along with strain and defects, to be investigated. The mass production of graphene is still in its infancy, and significant research is being conducted into growing large regions of pristine single-layer material. A contribution of Raman spectroscopy to this research will be discussed here.
Properties of graphene and graphene research
Graphene is a revolutionary two-dimensional material consisting of a single layer of carbon atoms. Electrons in graphene are able to travel ballistically, making it the most conductive material known to science. As a result, it can be foreseen that graphene will be the building block of the next generation of electronic devices.
Graphene is still a newly discovered material and there are many challenges that must be overcome before it can be used commercially. Its high conductivity relies on the material consisting of just a single layer of carbon atoms. As the number of layers increases, the properties of the graphene degrade until the point at which they match those of graphite. The production of large-area, single-layer graphene still requires significant research and development. Recently, the European Union set up a one billion euro research initiative focusing on graphene.1
Figure 1 – inVia Raman microscope. (Figures 1–4 courtesy of Renishaw plc.)
This is clearly an important area of research and investigation. An effective tool for studying the properties of graphene is the inVia Raman microscope from Renishaw plc (Old Town, Wotton-under-Edge, Gloucestershire, U.K.). See Figure 1.
The first Raman measurements on graphene were conducted by Andrea Ferrari et al. at the University of Cambridge in 2006.2 Using a Raman spectrometer from Renishaw, they were able to show that the Raman spectrum changes depending on the number of graphene layers present. This makes Raman an ideal tool for graphene research. Since then it has been found that Raman spectroscopy can be used to characterize a wide range of material properties in graphene, including stress and strain, doping, edges, electron mobility, disorder, defects, and thermal conductivity. An excellent review that discusses the capabilities of Raman on graphene can be found elsewhere.3
Raman spectroscopy as a tool for identifying the number of graphene layers
This article looks at how Raman spectroscopy can be used to examine the number of graphene layers present in flakes and films. Knowledge of this information is vital because the electrical properties of graphene decrease as the number of layers increases. Figure 2 illustrates Raman spectra collected from mechanically exfoliated flakes consisting of a single-layer, a bi-layer, and multiple layers of graphene. The number of layers was confirmed using atomic force microscopy (AFM) measurements. Here it can be seen that the 2-D Raman band undergoes a significant change in shape and also broadens with increasing number of graphene layers. This makes it an excellent indicator of the number of layers present in a flake.
Figure 2 – Raman spectra illustrating the 2-D band of graphene. These spectra were collected from mechanically exfoliated graphene flakes consisting of: a) a single layer of graphene, b) a bi-layer of graphene, and c) multiple layers of graphene. With increasing layer number, the 2-D band broadens and changes in shape.
Raman imaging of graphene
Renishaw’s patented StreamLine Plus fast imaging technique was used to collect a Raman image from mechanically exfoliated graphene flakes deposited on a SiO2/Si substrate. StreamLine differs from conventional Raman mapping in that it uses a laser line instead of a laser spot. This spreads the laser power over the sample, reducing the power density on the sample, and thus allowing ~20× more laser power to be used without causing any laser-induced modification of the sample. Because the Raman signal is proportional to the laser power used, measurement times are decreased by a factor of 20.
Conventional Raman mapping techniques superimpose an array of points on the sample area and take measurements at each point. This means that if the points do not overlap, areas of the sample are not measured. StreamLine can be operated in a mode called Slalom. In this mode, the laser line is zigzagged over the sample, ensuring that every part of the sample is measured and even small features contribute to a Raman spectrum. This makes StreamLine well-suited for conducting low spatial resolution survey work on samples of this type.
Figure 3 – a) White light image of graphene flakes dispersed on a 1 cm × 1 cm wafer. b) Corresponding composite StreamLine Raman image highlighting single-layer, bi-layer, and multilayer graphene flakes.
Raman spectra were collected from mechanically exfoliated graphene flakes dispersed on a ~1 cm × 1 cm SiO2/Si substrate. Because this area is much larger than the field of view of a 5× objective lens, a composite image or montage was created by moving the sample underneath the microscope and stitching the images together. WiRE spectrometer software (Renishaw) can be used to generate these images and conduct the stitching. The image, shown in Figure 3a, was used to define the Raman collection area. A StreamLine image was collected over a 10,490 μm × 10,724 μm area using a 20× objective lens and a 532-nm laser excitation source. In less than 90 min, 74,520 spectra were collected.
The data set was analyzed using a statistical technique—direct classical least squares (DCLS). The spectra of single-layer, bi-layer, and multilayer graphene shown in Figure 2 were used as references. This enabled each spectrum in the Raman data set to be identified as single-layer, bi-layer, or multilayer graphene. Figure 3b is a Raman image illustrating the distribution of graphene flakes of different thickness on the silicon wafer. Here it can be seen that the majority of the graphene flakes consist of multilayer graphene. Using particle statistics, it was possible to calculate the relative percentage of flakes that are single-layer, bi-layer, and multilayer as 2.2%, 5.7%, and 92.1%, respectively. It is also possible to calculate the average area of the single-layer and bi-layer flakes as 3944 μm2 and 4036 μm2.
Mechanical exfoliation is a common technique for making graphene flakes because of its low cost and ease of use. This technique may produce only a small percentage of graphene suitable for materials science research or for making test devices (single-layer and bi-layer graphene flakes). StreamLine Raman imaging has been demonstrated to be a valuable technique for surveying large areas of graphene flakes and locating these single-layer and bi-layer flakes. Since StreamLine can be applied to graphene made using any technique, it is an effective tool for graphene research. The technique has been also used to investigate the growth mechanism of graphene on copper to better understand how the orientation of the copper substrate affects the quality, thickness, and size of graphene films.4
High spatial resolution Raman imaging was conducted on a mechanically exfoliated graphene flake to determine its structure and properties. Figure 4 illustrates an optical image of the flake, along with the corresponding Raman images. As shown in Figure 4b, this flake contains regions of both a single-layer and bi-layer graphene. Also of interest is the G-band position, which varies significantly across the flake. In particular, it can be seen that the G-band is shifted to higher wave numbers in the single-layer regions. It has been shown elsewhere5 that the G-band of graphene shifts upwards with increasing electron concentration. This is in good agreement with the author’s layer assignment. Variations in the G-position within the layers can in part be explained by the change in electron concentration, but is also influenced by local strain and defects.6
Figure 4 – a) White light image of graphene flake; b) Raman image illustrating areas of single-layer and bi-layer graphene in green and white, respectively; c) Raman image illustrating change in graphene G-band position, which is an indicator of stress and doping.
StreamLine fast imaging is a powerful tool for surveying graphene. Using StreamLine, it was possible to characterize the size and distribution of single-layer, bi-layer, and multilayer regions of graphene over a 1-cm2 area. High-resolution Raman imaging allowed individual flakes to be characterized on a submicron-length scale. This enables the number of layers of graphene present to be determined, and local electron concentration, strain, and defects to be investigated.
- Ferrari, A.C.; Meyer, J.C. et al. Raman spectrum of graphene and graphene layers. Phys. Review Lett. 2006, 97(18), 187401.
- Ferrari, A.C.; Basko, D.M. Raman spectroscopy as a versatile tool for studying the properties of graphene. Nature Nanotechnol. 2013, 8(4), 235–46.
- Murdock, A.T.; Koos, A. et al. Controlling the orientation, edge geometry and thickness of chemical vapor deposition graphene. ACS Nano 2013, 7(2), 1351–9.
- Ferrari, A.C. Raman spectroscopy of graphene and graphite: disorder, electron–phonon coupling, doping and nonadiabatic effects. Solid State Commun. 2007, 143(1), 47–57.
- Mohiuddin, T.M.G.; Lombardo, A. et al. Uniaxial strain in graphene by Raman spectroscopy: G peak splitting, Grüneisen parameters, and sample orientation. Phys. Rev. B 2009, 79(20), 205433.
Dr. Tim Batten is an Application Scientist, Renishaw plc, Old Town, Wotton-under- Edge, Gloucestershire, GL12 7DW, U.K.; tel.: +44 (0) 1453 524 524; e-mail: [email protected].