Scholarly article on topic 'Effect of copper surface pre-treatment on the properties of CVD grown graphene'

Effect of copper surface pre-treatment on the properties of CVD grown graphene Academic research paper on "Nano-technology"

CC BY
0
0
Share paper
Academic journal
AIP Advances
OECD Field of science
Keywords
{""}

Academic research paper on topic "Effect of copper surface pre-treatment on the properties of CVD grown graphene"

Effect of copper surface pre-treatment on the properties of CVD grown graphene

Min-Sik Kim, Jeong-Min Woo, Dae-Myeong Geum, J. R. Rani, and Jae-Hyung Jang

Citation: AIP Advances 4, 127107 (2014); doi: 10.1063/1.4903369 View online: http://dx.doi.org/10.1063/1.4903369

View Table of Contents: http://scitation.aip.org/content/aip/journal/adva/4/12?ver=pdfcov Published by the AIP Publishing

Articles you may be interested in

Ultraviolet/ozone treatment to reduce metal-graphene contact resistance Appl. Phys. Lett. 102, 183110 (2013); 10.1063/1.4804643

Graphene based field effect transistor for the detection of ammonia J. Appl. Phys. 112, 064304 (2012); 10.1063/1.4752272

Impact of post-growth thermal annealing and environmental exposure on the unintentional doping of CVD graphene films

J. Vac. Sci. Technol. B 30, 041213 (2012); 10.1116/1.4731472

The effect of chemical residues on the physical and electrical properties of chemical vapor deposited graphene transferred to SiO2

Appl. Phys. Lett. 99, 122108 (2011); 10.1063/1.3643444

Multilayer epitaxial graphene formed by pyrolysis of polycrystalline silicon-carbide grown on c-plane sapphire substrates

Appl. Phys. Lett. 98, 132108 (2011); 10.1063/1.3575202

Eoodfellouj

metals • ceramics • polymers composites • compounds • glasses

Save 5% • Buy online

70,000 products • Fast shipping

(■) CrossMark

VHi «-dick for updates

Effect of copper surface pre-treatment on the properties of CVD grown graphene

Min-Sik Kim, Jeong-Min Woo, Dae-Myeong Geum, J. R. Rani, and Jae-Hyung Janga

Department of Information and Communications, Gwangju Institute of Science and Technology (GIST), Gwangju 500-712, Republic of Korea

(Received 29 September 2014; accepted 24 November 2014; published online 3 December 2014)

Here, we report the synthesis of high quality monolayer graphene on the pre-treated copper (Cu) foil by chemical vapor deposition method. The pre-treatment process, which consists of pre-annealing in a hydrogen ambient, followed by diluted nitric acid etching of Cu foil, helps in removing impurities. These impurities include native copper oxide and rolling lines that act as a nucleation center for multilayer graphene. Raman mapping of our graphene grown on pre-treated Cu foil primarily consisted of ~98% a monolayer graphene with as compared to 75 % for the graphene grown on untreated Cu foil. A high hydrogen flow rate during the pre-annealing process resulted in an increased I2D/IG ratio of graphene up to 3.55. Uniform monolayer graphene was obtained with a I2D/IG ratio and sheet resistance varying from 1.84 - 3.39 and 1110 - 1290 Q/m, respectively. © 2014 Author(s). All article content, except where otherwise noted, is licensed under a Creative Commons Attribution 3.0 Unported License. [http://dx.doi.org/10.1063/1.4903369]

I. INTRODUCTION

Graphene, a single-atom-thick sheet of hexagonally arrayed sp2-bonded carbon atoms, has received much attention due to its fascinating electrical, mechanical, optical, and thermal properties. These properties are derived from graphene's unique electronic band structure.12 For the high-volume manufacturing of electronic devices, the synthesis of large-area high-quality graphene is very important. Recently, various methods have been used to deposit graphene, such as mechanical micro-cleavage,3 reduction of graphene oxide,4 epitaxial growth on Silicon Carbide (SiC), and catalyst-assisted chemical vapor deposition (CVD).5,6 However, the mechanical exfoliation of graphite is a time-consuming process, especially when fabricating small-sized graphene. It is less efficient in addressing the need for mass fabrication of a large-area uniform monolayer graphene. Although the sublimation of SiC is a wafer-scale method, the SiC substrate is expensive and requires very high-temperatures. Therefore, it remains challenging to grow graphene at large scale with uniform thickness. Chemical vapor deposition (CVD), which grows large area graphene film onto transition metal substrates, is one of the most promising methods for the low-cost production of graphene film. The CVD method, which uses Cu or Nickel as the catalyst, has been proposed as a promising approach, offering higher efficiency, scalability, and large area deposition. The major advantage of CVD synthesized graphene is that following growth onto a catalytic metal surface, the resultant high quality single layer graphene films are readily transferable onto a multitude of substrates.

The CVD method requires precursors that contain carbons such as methane (CH4) or vapors of liquid carbon source like alcohols. Precursors pyrolyzed to carbon atoms at a high temperature can form a graphitic structure from dissociated carbon atoms. A large area graphitic structure can form at 2500°C without metal catalysts. Therefore, transition metals serve as efficient catalysts that transform hydrocarbons into graphitic materials at lower temperatures. Metal catalysts not only lower

aAuthor to whom correspondence should be addressed. Electronic mail: jjang@gist.ac.kr

2158-3226/2014/4(12)/127107/8 4,127107-1 ~ " iilli if J nil \

energy barriers for the pyrolysis of precursors, but also promote the growth of graphitic structure formation. There are various metal catalysts. Among the various metal catalysts (such as Ni, Cu, Pd, Ru, Ir etc), Ni, and Cu are particularly attractive because of their etchability and cost efficiency.7 Due to the different carbon solubility, Cu and Ni have different growth mechanisms, which were studied by Ruoff et al.8 Graphene grows on Ni through a dissolution-precipitation mechanism, while the graphene that grows on Cu requires a surface process. The growth on Cu suggests easier control over the deposition of graphene, so Cu has been regarded as a more viable substrate for obtaining monolayer graphene. However, there are several challenges to synthesizing graphene on Cu. Copper can be easily oxidized, so it results in a native oxide layer on the Cu surface, which hinders the growth of uniform graphene. In addition, Cu surface impurities may cause discontinuities and the growth of multilayer graphene. This is because impurities serve as a nucleation site for multilayer graphene, which degrades the uniformity of monolayer graphene.9 The availability of uniform graphene films will enable the reliable production of large arrays of identical graphene devices.

In this letter, we demonstrate the growth of high quality monolayer graphene on pre-treated Cu foil by CVD. The pre-treating process consists of H2 annealing, followed by acid (HNO3) cleaning. We observed that the H2 pre-annealing process eliminates native oxides, while nitric acid etches Cu, removing rolling lines and impurities. The Raman mapping of graphene grown on the pretreated Cu foils suggests that our graphene consists mostly of monolayer graphene with excellent uniformity. The graphene grown on pre-treated Cu foil has a higher I2D/IG ratio than that grown on untreated Cu foil. Our study suggests that the pre-treatment of Cu foil via H2 annealing and HNO3 treatment is a promising technique that can grow high quality and uniform monolayer graphene.

II. EXPERIMENTAL DETAILS

Graphene was synthesized on 25-^m-thick Cu foil (Alfa Aesar). The as-received Cu foil contains organic impurities and native copper oxide, which generally results in an undesirable growth of graphene. At first, diluted acetic acid, which has been demonstrated to partially remove Cu2O, was introduced to remove copper oxide.10 After rinsing with DI water, Cu foil was dipped in acetone for 15 minutes to remove organic impurities, followed by rinsing in methanol and DI water. After Cu foil cleaning, pre-annealing was conducted under a pressure of 100 Torr. Ar and H2 gas were able to flow by varying the H2 flow rate. After the annealing, the Cu foil was taken from the furnace and soaked in diluted nitric acid for various times.

The pre-treated Cu foil was loaded into the CVD chamber and the temperature was increased under Ar (495 SCCM) and H2 (5 SCCM) ambience. The Cu foil was kept at 1000°C for 30 minutes, and subsequently, CH4 (50 SCCM) was flowed into the chamber for 10 minutes. After that, the chamber was cooled down to room temperature under an Ar (495 SCCM) and H2 (5 SCCM) ambience. The graphene grown on one side of the Cu foil was dry etched by reactive ion etching. The polymer film of polymethyl methacrylate (PMMA), which was diluted in anisole (PMMA was purchased from Micro chem, No. M230004), was chosen to hold graphene. PMMA was poured onto the Cu foil and spin coated at a rotation speed of 3000 rpm for 40 seconds. The PMMA coated graphene on Cu foil was placed in 0.25 M of ammonium persulphate ((NH4)2S2O8) to etch the Cu foil for 12 hours. PMMA coated graphene was floated on the etchant and rinsed with DI water. PMMA coated graphene was scooped out from DI water onto a 300-nm SiO2/Si substrate. It was dried at room temperature and transferred to a hot acetone bath to remove PMMA. Subsequent steps of methanol and IPA cleaning were conducted. The Raman spectra and mapping were obtained using an inVia Raman Microscope system from Renishaw, Inc (excitation of 514 nm).

III. RESULTS AND DISCUSSION

Raman spectroscopy was used to probe the structural and electronic characteristics of graphite materials. It provided useful information on the defects (D band), in-plane vibration of sp2 carbon atoms (G band), as well as the stacking order (2D band). Raman spectroscopy is a powerful tool that characterizes the crystalline quality of graphene layers. The spectral features of graphene exhibited

a G peak at -1586 cm-1, a D peak at -1350 cm-1, and a 2D peak at -2686 cm-1.1112 The G and 2D peaks represent the E2g vibrational and out-of-plane modes within the aromatic carbon rings, respectively. The G band is a degenerated optical phonon mode at the Brillouin zone center and is induced by a single resonance process. It is extremely sensitive to strain effects and is also a good indicator of the number of graphene layers. As the number of layers increases, the G band position to lower frequencies. The ratio of intensities of 2D (I2D) and G (IG) band (I2D/IG ratio) is a criterion in determining the number of layers of synthesized graphene samples. AI2D/IG -2 indicates monolayer graphene, 2>I2D/IG > 1 indicates a bilayer, and I2D/IG < 1 indicates multilayers.13,14 The D peak is due to the breathing modes of six-atom rings and requires scattering at defect sites in order to conserve the momentum. It comes from transverse optical (TO) phonons around the K or K0 points in the first Brillouin zone, and involves an inter-valley double resonance process. However, double resonance can also happen in the intra-valley process, where it connects two points belonging to the same cone around K or K0. Thus, the intra-valley double resonance process has a D' peak at -1620 cm-1. Thus, D and D' peaks are a measure of the defects in the film. Also, the intensity ratio of the D and G peak (ID/IG) values are a measure of the disorder or defect and an increase in the ID/IG ratio indicates an increase in the disorder of the film.

The Raman spectroscopy and mapping of graphene grown on untreated Cu foil and transferred to SiO2/Si substrate is shown in Figure 1. The optical microscope image of the mapping area is shown in Figure 1(a). Figure 1(b) shows a representative Raman map showing the intensity ratio of 2D and G bands (I2D/IG) measured in a 50 ^m x 50 ^m area (corresponding to the optical microscope image shown in Figure 1(a)). Figure 1(c) shows the histogram of the I2D/IG ratio in the Raman map shown in Figure 1(b). Figure 1(d) depicts Raman spectra measured from spots in Figure 1(a).

Figure 1(b) shows that the majority of I2D/IG ratio images has two major colors. One is near 1.9 and the other is near 2.4. The average I2D/IG ratio is 2.1. However, the I2D/IG ratio appears in a range of

FIG. 1. The Raman spectroscopy and mapping of graphene grown on untreated Cu foil and transferred to SiO2/Si substrate, (a) optical microscope image of the mapping area, (b) representative Raman map showing the intensity ratio of 2D and G bands (I2D/IG) measured in a 50 ^m x 50 ^m area (corresponding to optical microscope image shown in (a)), (c) histogram of I2D/IG ratio in the Raman map shown in (b), (d) Raman spectra measured from a spot in (a).

FIG. 2. I2D/IG ratios variation depending on H2 flow rate during pre-annealing. Raman spectrum of the CVD graphene on SiO2/Si substrate at two different spots in the inset. (Inset) Optical image of the graphene showing a monolayer (black circle) and few-layer graphene (red circle). Scale bar is 10 ^m in the inset.

0.9 to 3.0. Only 90 points among the 121 points show a monolayer property where the I2D/IG ratio is greater than 2. Additionally, the average FWHM value is 43 cm-1 and 2D peak FWHM varies between 34 cm-1 - 60 cm-1. Thus, Raman mapping of graphene grown on untreated Cu foil suggests that nearly 75% of the area mapped exhibits a monolayer property with I2D/IG >2.

The Raman spectra (Figure 1(d)) shows a G peak at 1584 cm-1, a D peak around 1343 cm-1, and a 2D peak around 2686 cm-1, which is consistent with the previously reported values.11,12 The ratio of D to G peak intensities (ID/IG ratio) is found to be around 0.29, which suggests that there are some defects in the synthesized graphene. The D' peak is observed at 1620 cm-1, which also indicates disorder in the film.

Figure 2(a) shows the I2D/IG ratio of graphene grown on Cu foil annealed at various H2 flow rates. A pre-annealing condition of 5 SCCM H2 yields an average value of 2.15, but the range varies from 0.94 - 3.01. This shows that there was variation in terms of the number of graphene layers. The highest

I2D/IG, 3.55, was obtained at a H2 flow rate of 50 SCCM. However, a considerable amount of bilayer graphene with I2D/IG ~ 1 exists in the area. It is worth noting that bilayer graphene characteristics were only found on the graphene grown on periodic rolling lines of Cu foil. This indicates that rolling lines form nucleation sites to grow multilayer graphene. The synthesized graphene was transferred to the SiO2/Si substrate for characterization. The corresponding optical image of graphene grown on rolling lines, as well as that grown on the normal area of Cu foil with corresponding Raman spectrums, are shown in Figure 2(b).

The inset to Figure 2 shows an optical image that has color differences between graphene grown in the normal area and rolling lines. The transparent region exhibits a normal area, while the dark color shows a rolling line area. The interpretation is confirmed by the Raman spectra obtained at two different spots, which correspond to the colored circles in the inset of Figure 2(b). The calculated I2D/IG and FWHM in the dark area, which are represented by the red circle, are 1.4 and 48 cm-1, respectively. For the monolayer represented by the black circle, the I2D/IG ratio and FWHM are observed to be 3.47 and 39 cm-1, respectively, in Figure 2(b). Rough rolling lines appear when the Cu foil gets rolled. In spite of the long time needed for annealing, rough lines still exist, and they act as the nucle-ation center of multilayer graphene.

In order to remove these rolling lines from the Cu foil, after pre-annealing, we conducted Cu surface treatment with diluted nitric acid by varying the treatment times to remove rough rolling lines from the Cu surface. Nitric acid etching will remove impurities such as C, In, and Fe, which also act as nucleation sites that cause discontinuities and the growth of multilayer graphene.9 Previous reports suggest that while pre-cleaning using nitric acid, NO2 bubbles were generated, thereby effectively preventing the impurities from being re-deposited back onto the Cu surface.15 In comparison with the pre-cleaning method, the etching process after annealing is more effective in removing the impurities and rolling lines due to the absence of surface copper oxide. Surface copper oxide hinders the reaction

FIG. 3. The Raman spectroscopy and mapping of graphene grown on pre-annealed and one minute nitric acid treated Cu foil (transferred to SiO2/Si substrate), (a) optical microscope image of the mapping area, (b) representative Raman map showing the intensity ratio of 2D and G bands (I2D/IG) measured in a 50 ^mx 50 ^m area (corresponding to optical microscope image shown in (a), (c) histogram of the I2D/IG ratio in the Raman map shown in (b), (d) Raman spectra measured from a spot in (a).

between Cu and nitric acid. Through the slight copper etching using HNO3, impurities and rolling lines, which existed on the Cu foil, were removed.

The Raman spectroscopy and mapping of graphene grown on pre-annealed and HNO3 treated Cu foil, and transferred to the SiO2/Si substrate, is shown in Figure 3. The optical microscope image of the mapping area is shown in Figure 3(a). Figure 3(b) shows a representative Raman map indicating the I2D/IG ratio measured in a 50 ^m x 50 ^m area (corresponding to the optical microscope image shown in Figure 3(a)). Figure 3(c) shows the histogram of the I2D/IG ratio in the Raman map shown in Figure 3(b). Figure 3(d) depicts the Raman spectra measured from spots in Figure 3(a). From the figure, we can see the effect of pre-annealing and nitric acid treatment on the uniformity and number of synthesized graphene layers by comparing with that grown on untreated Cu foil.

The I2D/IG ratio image mapping area also shows two major colors. One is near 2.5 and the other is near 3.3. The average and range of the I2D/IG ratio is 2.6 and 1.8 - 3.4. Among the 121 points selected, 119 points exhibit monolayer properties. The average value and range of 2D peak FWHM are 38 cm-1 and 35 cm-1 - 48 cm-1, respectively. From Figure 3(d), the spectra shows a G peak at 1586 cm-1, a D peak around 1343 cm-1, and a 2D peak around 2686 cm-1. The 2D band of Raman spectrum (Figure 3(d)) can be fitted by a single Lorentzian with a center at ~2686 cm-1 and FWHM ~35 cm-1.

Graphene grown on pre-annealed and nitric acid treated Cu has a higher I2D/IG ratio and narrow 2D peak FWHM, which indicates uniformity in the number of graphene layers. Based on our I2D/IG data, we conclude that most of the area mapped is monolayer graphene, 98.3% of the area having a monolayer property (I2D/IG >2). The average ID/IG is found to be 0.103, which suggests a very low defect density in these films. The disorder-induced D band in the spectra shown in Figure 3(d) is observed to be very small, indicating the high quality of the graphene. Also, the absence of a D' peak also suggests low disorder in the film. Thus, the graphene grown on the pretreated Cu surface was of high quality, as indicated by a high I2D/IG ratio and the absence of a disorder-induced D' band in the Raman spectra.

Figure 4 shows the I2D/IG ratio and sheet resistance of synthesized graphene at various nitric acid treatment times.

The highest average I2D/IG ratio obtained for the 1 minute treated Cu foil was 2.66. In terms of sheet resistance, graphene synthesized on 1 minute treated Cu foil also showed the lowest value in the

FIG. 4. I2D/IG ratios and sheet resistances of graphene by varying the nitric acid treatment with error bar.

TABLE I. The reported Sheet resistance of graphene deposited using various methods.

Deposition method Sheet resistance (Q/d) Reference

CVD CVD CVD CVD CVD

Solution processed RGO

Chemically and thermally reduced solution processed GO Chemical reduced exfoliated graphene oxide (GO)

range of 1110 to 1290 Q/m. For comparison, reported sheet resistance values of the graphene sheets synthesized using CVD and other methods are shown in Table I.

However, the sheet resistance dramatically increased when treatment was prolonged to 3 minutes. This was due to the severe damage of the Cu foil during the treatment. The sheet resistance of graphene will be increased, when there exists considerable cracks in the transferred graphene film. The rolling lines on the Cu foil will result in a film with cracks. When compared to graphene deposited on untreated Cu foil, the observed sheet resistance was higher for that deposited on Cu foil treated with HNO3 for 20 seconds. We believe that 20 seconds treatment was too short and it would have resulted in uneven and incomplete removal of rolling lines from the foil, which in turn increased the cracks on the deposited graphene, when compared to untreated Cu foil. We found that HNO3 treatment for 60 seconds to be optimum for removing rolling lines. Prolonged treatment after 60 seconds, etches the Cu foil, thus introducing many cracks in the deposited graphene and further increase in the sheet resistance.

IV. CONCLUSION

In summary, we report the effect of pre-H2 annealing and nitric acid treatment to synthesize high quality and uniform graphene on Cu. The highest I2D/IG, 3.55, was obtained at a H2 flow rate of 50 SCCM with a considerable amount of bilayer graphene. This was due to Cu surface impurities and rolling lines. Cu surface impurities and rolling lines were removed by using diluted nitric acid after pre-annealing the Cu foil. Graphene grown on Cu foil with pre-annealing and nitric acid treatment exhibited a high I2D/IG ratio that varied from 1.84 - 3.39. Raman mapping of graphene grown on pretreated Cu foil indicated that 98.3% of the mapped area exhibited monolayer characteristics with I2D/IG >2 compared to 75% monolayer characteristics for that grown on untreated Cu foil. The average ID/IG of graphene grown on pre-treated Cu foil was found to be 0.103 compared to 0.290 for that grown on untreated Cu foil. Graphene synthesized on 1 minute treated Cu foil showed the sheet resistance value in the range of 1110 to 1290 Q/m. Our results suggest that CVD graphene films grown on pre-treated Cu foil via H2 annealing and HNO3 etching have excellent quality and uniformity and consist mainly of a monolayer with a very low defect density.

ACKNOWLEDGMENT

This work was supported by the National Research Foundation of Korea through the grant (No. 2011017603)

1 A. K. Geim and K. S. Novoselov, Nature Mater. 6, 183 (2007).

2 K. S. Kim, Y. Zhao, H. Jang, S. Y. Lee, J. M. Kim, K. S. Kim, J.-H. Ahn, P. Kim, J.-Y. Choi, and B. H. Hong, Nature 457, 706 (2009).

3 K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva, and A. A. Firsov, Science 306, 666 (2004).

4 S. Stankovich, D. A. Dikin, R. D. Piner, K. A. Kohlhaas, A. Kleinhammes, Y. Jia, Y. Wu, S. T. Nguyen, and R. S. Ruoff, Carbon 45, 1558 (2007).

5 P. W. Sutter, J.-I. Flege, and E. A. Sutter, Nature Mater. 7, 406 (2008).

200-2100 700 125 2000-9000 1110-2300 100-1000 2000 1000-9000

This work

20 21 22

6 J. Coraux, A. T. N'Diaye, C. Busse, and T. Michely, Nano Lett. 8, 565 (2008).

7 X. Li, W. Cai, J. An, S. Kim, J. Nah, D. Yang, R. Piner, A. Velamakanni, I. Jung, E. Tutuc, S. K. Banerjee, L. Colombo, and R. S. Ruoff, Science 324, 1312 (2009).

8 X. Li, W. Cai, L. Colombo, and R. S. Ruoff, Nano Lett. 9, 4268 (2009).

9 J. W. Suk, A. Kitt, C. W. Magnuson, Y. Hao, S. Ahmed, J. An, A. K. Swan, B. B. Goldberg, and R. S. Ruoff, ACS Nano 5, 6916(2011).

10 K. L. Chavez and D. W. Hess, J. Electrochem. Soc. 148, G640 (2001).

11 Z. Ni, Y. Wang, T. Yu, and Z. Shen, Nano. Res. 1, 273 (2008).

12 J. R. Rani, J. Oh, J.-e. Park, J. Lim, B. Park, K. Kim, S.-J. Kim, and S. Chan Jun, Nanoscale 5, 5620 (2013).

13 A. C. Ferrari, Solid State Commun. 143, 47 (2007).

14 L. M. Malard, M. A. Pimenta, G. Dresselhaus, and M. S. Dresselhaus, Phys. Rep. 473, 51 (2009).

15 S. M. Kim, A. Hsu, Y. H. Lee, M. Dresselhaus, T. Palacios, K. K. Kim, and J. Kong, Nanotechnol. 24, 365602 (2013).

16 X. Li, Y. Zhu, W. Cai, M. Borysiak, B. Han, D. Chen, R. D. Piner, L. Colombo, and R. S. Ruoff, Nano Lett. 9, 4359 (2009).

17 A. Reina, X. Jia, J. Ho, D. Nezich, H. Son, V. Bulovic, M. S. Dresselhaus, and J. Kong, Nano Lett. 9, 30 (2009).

18 S. Bae, H. Kim, Y. Lee, X. Xu, J.-S. Park, Y. Zheng, J. Balakrishnan, T. Lei, H. Ri Kim, Y. I. Song, Y.-J. Kim, K. S. Kim, B. Ozyilmaz, J.-H. Ahn, B. H. Hong, and S. Iijima, Nat. Nanotechnol. 5, 574 (2010).

19 G. V. Bianco, M. Losurdo, M. M. Giangregorio, P. Capezzuto, and G. Bruno, Physical chemistry chemical physics: PCCP 16, 3632 (2014).

20 H. A. Becerril, J. Mao, Z. Liu, R. M. Stoltenberg, Z. Bao, and Y. Chen, ACS Nano 2, 463 (2008).

21 S. J. Wang, Y. Geng, Q. Zheng, and J.-K. Kim, Carbon 48, 1815 (2010).

22 J. Geng, L. Liu, S. B. Yang, S.-C. Youn, D. W. Kim, J.-S. Lee, J.-K. Choi, and H.-T. Jung, The Journal of Physical Chemistry C 114, 14433 (2010).