Studies of the Chemical Vapor Deposition Method of Generating Graphene
CHAPTER ONE
OBJECTIVE OF STUDY
The graphene growth by CVD process as has been seen as a viable technique for the fabrication of large–area SLG.
In this work, we set to explore the graphene growth in a cold–wall CVD system and the possibility of transferring its optimized growth recipe to the most common hot–wall tubular reactor.
CHAPTER TWO
THEORETICAL REVIEW
Graphene
One of the most recent advances in material science is the discovery of the so called carbon-based nanomaterials. The key to these kinds of systems is the so called sp2 hybridization. Carbon has the atomic number six, meaning that it has two core electrons occupying the 1s orbitals which are closely bound to the nucleus and doesn’t participate in atomic bonding. Carbon also has four valence electrons in the 2s and 2p orbitals. These electrons are more flexible and are able to take part in atomic bonding, such as C-C binding. When carbon atoms bind to each other, these orbitals can mix into spn hybridized orbitals, where n = 1, 2, 3.
In the case of sp2 systems, one 2s and two 2p orbitals mix to create three in-plane bonds. One example of such a system is the C60 molecule which was discovered in 1985. It is just one molecule of 60 carbon atoms in a spherical structure. Another example is the cabon nanotube. The simplest sp2 system, however, is called graphene, a one atom thick hexagonal grid of carbon atoms.
CHAPTER THREE
METHODOLOGY
Graphene synthesis
Pristine graphene synthesis
A schematic figure of the experimental setup, provided by the physics department, can be seen in Figure 3.1.
Figure 3.1: A schematic picture of the experimental setup for the graphene synthesis. The various gases enter the quartz tube from the right in the figure and is led through the CVD oven (Nabertherm GmbH) where the copper substrate is placed. Then the gases exit the oven and passes a pressure sensor (Leybold Vakuum Ionivac) and thereafter exit the system through the vacuum pump (Edwards).
The copper used for this thesis work was 127µm thick annealed copper foil (99.9%) from Alpha Aesar since the purity was proven to be sufficient and it was thick enough not to move around in the quartz tube due to the gas flow. Before the CVD process, the copper substrate was cleaned in glacial ascetic acid for 15 minutes to remove potential copper oxide since it reduces the catalytic activity [7]. It was then rinsed with ethanol and deionized water and dried it with a nitrogen gun and immediately placed inside the quartz tube.
CHAPTER FIVE
RESULTS AND DISCUSSION
Pristine graphene
The first sample that was synthesized was pristine graphene, as described in Section 3.1. The recipe used can be seen in Table 3.1. This section is divided into two parts, one about the as grown graphene on copper and the other about the graphene transferred to a target substrate called Flourine doped Tin Oxide (FTO). FTO is a transparent conducting film most commonly used in photovoltaic applications.
Pristine graphene on copper
An SEM picture of the graphene on the copper substrate can be seen in Figure 4.1. There are four areas of interest in this picture, marked 1 through 4 in the figure. The first area is most probably one layer of graphene, though it’s hard to say with only SEM. But if that is the case, areas 2 through 4 are probably two, three and four layers of graphene, respectively. But as mentioned in previous sections, an SEM picture is not a definitive answer to if the sample truly contain graphene nor its quality.
A better way to determine if area 1 is graphene is to use Raman spectroscopy. A heatmap of the ratio between the height of the 2D and G peaks can be seen in Figure 4.2. As explained in Section 2.3.2, this ratio is an indication of the number of layers of graphene in a sample, and according to the values in Figure 4.2 there are most likely two layers except for a smaller area of one layer graphene in the top left corner. The mean Raman spectrum of this area can also be seen in Figure 4.3 which also suggest that there are most likely two and one layers of graphene in the area. The D-band in this sample is also quite low as compared to the G-band, indicating that the graphene is of high quality and a low number of defects.
CHAPTER FIVE
SUMMARY AND CONCLUSION
The goal of this thesis study was to develop a CVD based method for graphene growth on copper with acetylene as precursor. The effects of ammonia treatment during growth was also investigated. A method for transferring the grown graphene sheets from the copper substrate with PMMA to another target surface, in this case FTO, was also to be developed to be able to use the graphene in applications. The characterization techniques used for investigating the grown graphene samples were Raman spectroscopy, scanning electron microscopy and x-ray photoelectron spectroscopy. With the CVD based method, three different samples of graphene were synthesized, one pristine and two ammonia treated ones with different growth times.
All the measurements made on the pristine graphene on the copper substrate indicated that the foil was covered by one- and few layer graphene with high quality. During the transfer to FTO, the graphene got tared by mechanical strains. There was also a high number of microscopic tares in the graphene sheet, probably caused by the thermal expansion of the PMMA compared to the graphene. But other than these tares, the transferred graphene was of high quality and it also seemed like it was only one layer, suggesting that only the graphene layers in contact with the PMMA was successfully transferred to the FTO.
The ammonia treated sample with the same growth time as the pristine graphene was then synthesized. According to the Raman spectra and the SEM pictures, there seemed to be fewer layers. This might be because of the lower relative concentration of precursor during the growth. There were also a large number of salt-like structures present before the transfer, and according to Raman it was probably cupric oxide. The XPS showed no sign of nitrogen above the noise level, meaning that the doping attempt was a failure. There was also a doublet in the XPS spectra suggesting that the sample contained antimony. Otherwise, the number of defects was very low and the graphene was of high quality. After the transfer, there was no trace of the anomalies meaning tat they were probably corroded by the ferric chloride, further suggesting that the anomalies were cupric oxide and not of organic origin. There were also some larger and microscopic tares in the graphene. But the quality of the transferred graphene was overall of high quality with a low number of defects.
The last graphene sample synthesized was ammonia treated with a longer growth time. The number of layers of this graphene was higher than both the previous samples, meaning that the longer growth time results in more layers. There were also some trace of the anomalies similar to the ones in the last ammonia treated sample, but way fewer. Otherwise the number of crystallographic defects was low and the quality of the graphene was overall high. There was no trace of nitrogen in this sample either, meaning that the doping attempt was a failure in this case as well. There was also a higher amount of antimony present. After the transfer, the number of layers was way lower, further suggesting that only the layers in contact with the PMMA gets transferred to the FTO. There were also fewer tares in this sample, possibly meaning that the ammonia treatment has some role in the manageability of the graphene. Overall the quality was high with few defects.
To summarize, this thesis work was a good, broad groundwork for future studies. It was proven that graphene growth and transfer was entirely possible with the setup provided. To get a deeper understanding of the different growth parameters, more work needs to be done. Some ideas could be to vary only one or two parameters, such as pre-treatments, annealing time, growth time, precursor concentration and pressure. Even other types of metal substrates could be tested, such as Ni, or a different precursor, like methane or hexane. There are also much improvement to be done in the transfer process since it is quite crude. The first thing would probably be to reduce the number tares.
REFERENCES
- Nobel Media Press release: The 1930 nobel prize in physics, June 2017.
- Nobel Media Press release: The 1981 nobel prize in physics, June 2017.
- Nobel Media Press release: The 1986 nobel prize in physics, May 2017.
- Nobel Media Press release: The 2010 nobel prize in physics, March 2017.
- Riichiro Saito Ado Jorio, Mildred Dresselhaus and Gene F. Dresselhaus. Raman Spectroscopy in Graphene Related Systems. WILEY-VCH Verlag,
- Beams, L. G. Can¸cado, and L. Novotny. Raman characterization of defects and dopants in graphene. J. Phys.: Condens. Matter, 27, 2015.
- L. Chavez and D. W. J. Hess. A novel method of etching copper oxide using acetic acid. Electrochem. Soc., 148, 2001.
- Debbichi, M. C. Marco de Lucas, J. F. Pierson, and P. Kru¨ger. Vibrational Properties of CuO and Cu ¡sub¿4¡/sub¿ O ¡sub¿3¡/sub¿ from First-Principles Calculations, and Raman and Infrared Spectroscopy. The Journal of Physical Chemistry C, 2012.
