- Email: [email protected]
Large area coating of graphene at low temperature using a roll-to-roll microwave plasma chemical vapor deposition
Thin Solid Films 532 (2013) 89–93
Contents lists available at SciVerse ScienceDirect
Thin Solid Films journal homepage: www.elsevier.com/locate/tsf
Large area coating of graphene at low temperature using a roll-to-roll microwave plasma chemical vapor deposition Takatoshi Yamada ⁎, Masatou Ishihara, Masataka Hasegawa Nanotube Research Center, National Institute of Advanced Industrial Science and Technology (AIST), 1-1-1 Higashi, Tsukuba 305-8565, Japan Technology Research Association for Single Wall Carbon Nanotubes (TASC), 1-1-1 Higashi, Tsukuba 305-8565, Japan
a r t i c l e
i n f o
Available online 6 January 2013 Keywords: Graphene Plasma CVD Roll-to-roll deposition Coating Raman spectroscopy
a b s t r a c t Roll-to-roll microwave plasma chemical vapor deposition (CVD) at low temperature has been developed for the fabrication of a continuous and large area deposition of graphene films for the application of transparent conductive films. Deposition of graphene films on copper (Cu) foils, under methane (CH4) and hydrogen (H2) plasma below 380 °C, were confirmed and characterized by Raman spectroscopy. In addition, the film qualities were improved by controlling the CH4/(CH4 + H2) ratio and the process pressure. The graphene film obtained by roll-to-roll microwave plasma CVD process exhibited uniform Raman spectra towards the width direction of the Cu foil (A4 width). Transmittance and haze of the films were confirmed to be uniform and are in sufficient quality for the use of practical touch panel applications. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Transparent conductive films are essential components for many modern electronic and optoelectronic applications such as displays, touch screens, solar cells and light emitting devices. Although the commercial transparent conductive films used for practical applications are made from indium tin oxide (In2O3:Sn, ITO), a substitute material for ITO has been a topic of great interest, due to the limited supply of indium and other rare-metals. In addition, recent advance in organic semiconductor have realized flexible electro-optics devices such as electronic papers and sheet-type solar cells, for which inflexible ITO films are not applicable. Therefore, development of an alternative transparent conductive film has attracted much attention. Graphene, which consists of only carbon (C) atoms, is expected to be one of the most appropriate materials for transparent conductive films, since an ideal mono-layer graphene has a transmittance of 97.7%, an electron mobility of 200,000 cm 2/Vs at room temperature and a thermal conductivity of 5000 W/mK. However, the development of graphene mass production technology has not been established up to now, since it requires either high temperature or long process time. Techniques such as thermal chemical vapor deposition (CVD) on metal catalysis require high temperature (about 1000 °C) and long process time (few tens of minutes) to form graphene films [1]. As for low
⁎ Corresponding author at: Nanotube Research Center, National Institute of Advanced Industrial Science and Technology (AIST), 1-1-1 Higashi, Tsukuba 305-8565, Japan. E-mail address: [email protected] (T. Yamada). 0040-6090/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.tsf.2012.12.102
temperature formation of graphene film, reduction process of graphene oxide was reported [2]. Although this process enables us to form large area graphene on various substrates using spin coating technique, damages and defects during oxidization process still remain in the formed graphene films. Using surface wave plasma CVD, high quality graphene films were obtained at low temperature and few tens of seconds [3]. In addition, touch panel application based on surface wave plasma CVD graphene was demonstrated [3]. From the previous reports, it is expected that the graphene deposition process using surface wave plasma CVD is suitable for an industrial mass production. Roll-to-roll process is another common technique for industrial mass productions of large-area thin film. By combining roll-to-roll process and CVD technique, a continuous deposition process of graphene film is expected. Among various CVD methods, surface wave plasma CVD is selected for combination with roll-to-roll process, because of its possible low-temperature deposition of graphene. In this paper, we describe the development of roll-to-roll microwave plasma (MWP) CVD process towards the industrial mass production of graphene films. In our previous report [3], Ar was used as the carrier gas to keep the plasma stable. However, Ar was considered to damage the graphene and incorporates unintentional impurities from the CVD system. Therefore, effects of flow ratio of CH4 to the total gas (CH4 + H2) and process pressure during CVD process on graphene qualities were examined in this study. A continuous graphene film with 294 mm width is deposited on copper (Cu) foils at low temperature of around 380 °C. High uniformity of transparency (95%) in large area is confirmed. Our data indicated that the roll-to-roll MWPCVD was a promising technique for the graphene mass production towards the transparent conductive films.
90
T. Yamada et al. / Thin Solid Films 532 (2013) 89–93
Fig. 1. The schematic of a roll-to-roll microwave plasma CVD apparatus based on a linear antenna type.
2. Experimental details 2.1. A roll-to-roll system in a linear antenna type microwave plasma CVD for graphene deposition A MWPCVD process based on a linear antenna type system has been one of the potential technique for graphene deposition over large area at low process temperature, because MWP sustained by surface wave mode generates high plasma density (10 11–10 12 cm−3) even at relatively low pressure (b 100 Pa), which is effective to keep the substrate at low temperature. Large area coating of nano-structured diamond films was established using such CVD processes [4,5]. Our linear antenna type MWPCVD system (Fig. 1) had eight coaxial linear antennas covered with quartz tubes, which were used to generate the plasma. The antennas were cooled by air in quartz tubes. The CVD apparatus was equipped with two microwave generators of 2.45 GHz with a maximum power of 20 kW. Source gases (methane: CH4 and hydrogen: H2) were introduced from the top of CVD chamber and evacuated from the bottom of the chamber using a rotary pump [6]. The roll-to-roll system consists of a pair of winder and unwinder. The winder has a motor attached in order to control the flow speed of metal foils in the range from 1 to 500 mm/s. The unwinder is equipped with a brake works to keep the appropriate tension of the metal foils. The sample holder is cooled by water supply. The metal foils are exposed to plasma during traveling of 48 cm length of plasma area. The
Fig. 3. The peak intensity ratio as a function of the CH4/(CH4 + H2) ratio.
numbers of deposited graphene layers can be controlled by foil transfer speed. A rolled 33 μm-thick Cu foils with 294 mm in width, prepared by rolling processes, are used as the substrates in this study. A metal foil of up to 30 m in length and 33 μm in thickness could be installed in our unwinder section [7]. The typical CVD conditions in order to form continuous films were as follows. The microwave power was 12 kW. The temperature of sample stage was less than 380 °C. The flow ratio of CH4 to the total gas (CH4 +H2) [CH4/(CH4 +H2)] and the process pressure was controlled as parameters in order to obtain large domain size. The CH4/(CH4 +H2) ratios were changed from 10 to 70%. In order to keep plasma stably, Ar was used during CVD process in our previous report [7]. However, Ar was not added in this study since Ar was considered to make damage on graphene and to make unintentional impurities from CVD and roll-to-roll system. The pressures in our CVD system were controlled from 30 to 300 Pa. The Cu foil follow speed was 5 mm/s, which let the Cu foil exposed to plasma for 96 s.
2.2. Characterization techniques of graphene films
Fig. 2. Raman spectra of the obtained films, where the CH4/(CH4 + H2) ratios were 10, 30, 50 and 70%.
The obtained graphene films on Cu foils were characterized by Raman spectroscopy excited by a 638 nm diode laser with a beam spot size of 1 μm in diameter. Plane view transmission electron microscopy (TEM) observations were also carried out. Graphene films were mounted on Cu mesh grid. The accelerated voltage was 200 kV. Cross-sectional transmission electron microscopy (XTEM) observations were performed to estimate the length of graphene and the number of layers. In order to prevent unnecessary damage of graphene film prior to the XTEM observation, amorphous carbon was deposited on top of the graphene/Cu substrate. Gallium (Ga) focus ion beam (FIB) etching technique was used for sample preparations. For observations, accelerated voltage was 300 keV and the resolution of TEM used in this study was about 2 nm. In order to evaluate a transmittance, haze and sheet resistance of the film, the substrates (Cu foils) were removed by dipping the sample into a solution of ferric chloride (FeCl3) and the graphene films were transferred onto PET films. The transmittance and the haze at 550 nm were measured by means of visible light spectroscopy. The illuminated area was about 0.7 mm in diameter. Four-probe method was used to measure the sheet resistance. The distance between probes was 1 mm. The measurements were carried out in air at room temperature.
T. Yamada et al. / Thin Solid Films 532 (2013) 89–93
91
Fig. 5. XTEM image of the obtained graphene where the CH4/(CH4 + H2) ratio was 50% and the pressure was 30 Pa.
deposited under the condition where the CH4/(CH4 + H2) ratios were less than 50%, while graphitic carbon film was deposited when the CH4/(CH4 + H2) ratios was 70% [6,8]. As CH4/(CH4 + H2) ratio increases, the number of carbon radicals that reach Cu surface increases. It is speculated that graphitic components formation is dominant rather than graphene formation under the conditions of high carbon radical atmosphere. In order to discuss the quality of the obtained graphene, the D-band to G-band peak intensity ratios (ID/IG) were evaluated, since the ID/IG ratios are related with the average domain size (La) by the following equation [8]: −10 4 −1 La ¼ 2:4 10 λ ðID =IG Þ
ð1Þ
where λ is wave length of laser (638 nm). As increasing the CH4/ (CH4 + H2) ratio, the ID/IG ratio decreases. This indicates that the maximum domain size of the graphene film was obtained by the CVD condition where the CH4/(CH4 + H2) was 50%. The estimated domain size was almost the same as the reported size obtained by plasma CVD [7,9]. Although the lines due to graphite lattice patterns in plane view TEM image as shown in Fig. 4 (a), atomic structures and flake boundaries are not observed, fast Fourier transform (FFT) pattern [Fig. 4 (b)] obtained from Fig. 4 (a) shows hexagonal diffraction pattern and ring pattern, which indicates that the obtained film is graphene. However, the domain length estimated from the XTEM image (as shown in Fig. 5) was in the range of 5 to 15 nm. This is much smaller than that estimated from Raman results. It was reported that the contribution to the peak intensities of D-band from the armchair edge was larger than from the zigzag edges [10]. Therefore, we have to carefully estimate the flake size using Raman results. Although Raman spectroscopy is less powerful to determine the domain length, it enables one to simultaneously evaluate film quality of graphene by using ID/IG ratio, since D-band is attributed to defects. From Raman spectra, we concluded the CVD condition where Fig. 4. (a) Plane view TEM image of the obtained graphene where the CH4/(CH4 + H2) ratio was 50% and the pressure was 30 Pa. And (b) FFT pattern.
3. Results and discussion 3.1. Graphene deposition by linear antenna type microwave plasma CVD In order to investigate the depositing conditions of graphene films on Cu foils using the linear antenna type microwave plasma CVD without the roll-to-roll process, the CH4/(CH4 + H2) ratios and the process pressures were the adjusted parameters. Deposition time was fixed to 90 s for all samples. The films deposited by four kinds of the CH4/(CH4 + H2) ratios at 30 Pa were characterized by Raman spectroscopy (Fig. 2). All Raman spectra show G- and D-bands around 1350 and 1585 cm −1, respectively. Peaks due to 2D-band around 2690 cm −1 are obtained from the films prepared by the CVD conditions where the CH4/(CH4 + H2) ratios were 10, 30 and 50%. Peaks of D′-band are also detected in all spectra. Both of G and 2D-band are attributed to graphene, and D- and D′-bands are originated from defects and/or edges in graphene. By changing the CH4/(CH4 + H2) ratio, the film quality can be modified. The graphene films were
Fig. 6. Raman spectra of the films deposited under pressures at 300 Pa and 30 Pa.
92
T. Yamada et al. / Thin Solid Films 532 (2013) 89–93
Fig. 7. Raman spectra of graphene film prepared by roll-to-roll MWPCVD on Cu foil measured at 4-cm intervals.
the CH4/(CH4 +H2) was 50% is the best condition for our experiment. The 2D-band to G-band peak intensity ratios (I2D/IG) were evaluated to discuss the numbers of layers. From the literature [10], two or three layers were expected after 90 s of CVD process regardless of CH4/(CH4 +H2) concentration. Although it is considered that the deposition rate and size of graphene films increase with CH4/(CH4 +H2) concentration, we concluded CH4/(CH4 +H2) ratio of 50% is the best condition for our experiments, considering the film quality evaluated from Raman spectra. It is speculated that graphitic components formation is dominant rather than graphene formation under the conditions of high carbon radical atmosphere. It must be noted that no signal due to silicon was detected by energy dispersive X-ray spectroscopy (EDX), which will be discussed in elsewhere. The typical Raman spectra obtained from films prepared at 30 and 300 Pa, where the CH4/(CH4 + H2) ratio was 50%, are shown in Fig. 6. It is clear that both spectra show peaks due to G- and D-bands. However, only films obtained at 30 Pa have 2D peak. The Raman data indicate that graphene films were obtained at low process pressure (30 Pa), while graphitic films were deposited at high process pressure (500 Pa). So far, we have not clarified the radical species that are responsible for graphene formation in our plasma CVD. However, it is considered that the first layer, which covers the Cu surface in the beginning, forms graphene but the second layer and up do not form graphene at high process pressure. It is speculated that graphitic component formation is dominant rather than graphene formation under the conditions of high carbon radical atmosphere. From the Raman results as shown in Figs. 3 and 6, both of low CH4/(CH4 + H2) ratio and low process pressure are important to lower the peak intensity of D-band which is attributed to defects or edges.
(1588.2 cm−1), 2D-band (2661.8 cm−1), D-band (1330.8 cm−1) and D′-band (1625.0 cm−1) were observed. Using Eq. (1), the estimated average domain size was about 10 nm. Raman results indicate that the obtained graphene films synthesized by the roll-to-roll microwave plasma CVD consist of small size of domains. It must be noticed that no major differences in Raman spectra compared to our previous reports [3,7], which indicates that the obtained graphene films prepared by roll-to-roll microwave plasma CVD had similar structures with the previous reports [3,7]. The transmittance, the haze and the sheet resistance of the transferred graphene/PET film structures at an interval of 4 cm are shown in Fig. 8. The transmittance includes that of the PET substrates, which is about 91%. The total transmittance is about 89%, the haze is from 0.95 to 1.3%, and the sheet resistance is 0.9–3 × 106 Ω/sq. The obtained transmittance and the sheet resistance are almost the same as those of the reported graphene-based transparent conductive films obtained from the reduction process of graphene oxide [2]. However, the obtained sheet resistance was three or four orders higher than the data of graphene film deposited by thermal CVD. Therefore, the reduction of the sheet resistances is necessary to utilize the practical transparent conductive films in order to use transparent conductive graphene films for practical touch panel applications. The difference in the sheet resistance obtained by MWPCVD and by thermal CVD would be explained by the small crystal flake of the deposited graphene films in the MWPCVD at extremely low temperature compared with that of the thermal CVD, since in the case of the thermal CVD graphene such a high intensity D-band is usually not observed [1]. In addition, the reduction of the sheet resistance by the doping during transfer process was reported [1]. By optimizing the experimental conditions of the CVD process and the transfer process, the reduction of the sheet resistance would be expected. Using surface wave plasma CVD, graphene was deposited on aluminum (Al) foils, which has no catalytic effect to decompose hydrocarbon and hydrogen [3]. Therefore, we expect that graphene can directly be deposited on PET films. Although graphitic carbon films were deposited on PET film by surface wave plasma CVD, we have not yet confirmed graphene formation on PET films. By taking graphene deposition on
3.2. Graphene deposition by roll-to-roll microwave plasma CVD From the obtained results mentioned above, graphene films were deposited by the roll-to-roll microwave plasma CVD under the condition of 50% of the CH4/(CH4 + H2) ratio and 30 Pa of the pressure. Typical Raman spectra of the graphene film at interval of 4 cm in the across-the-width direction of the Cu foil are shown in Fig. 7, which suggests good uniformity of the film. The Raman signals of G-band
Fig. 8. Transmittances, hazes and sheet resistances of roll-to-roll MWPCVD graphene/PET structures measured at 4-cm intervals.
T. Yamada et al. / Thin Solid Films 532 (2013) 89–93
Cu foil into account, it would be expected that graphene deposition on PET film by optimizing the CVD conditions. Up-to-now, the catalytic effect of Cu plays an important role for the deposition of graphene with high rate and low temperature. This is under investigation. 4. Summary We developed a process by combination of the plasma CVD process at low substrate temperature and the roll-to-roll process for the graphene mass production. By selecting the microwave plasma sustained by surface-wave mode, high plasma density (10 11–1012 cm−3) even at relatively low pressure (b 100 Pa) was realized, which is effective to keep the low substrate temperature and to obtain the high deposition rate. It was found from Raman results that the qualities of graphene films were strongly depended on the CH4/(CH4 +H2) ratios and the process pressures. The graphene film obtained by the roll-to-roll microwave plasma CVD process showed uniform Raman spectra towards the width direction of the Cu foil (A4 width). The high uniform transmittance and haze were confirmed, which were enough for the practical touch panel application. By optimizing the conditions of both CVD and transfer process, the improvement of the electrical conductivity will be attained. Our experimental data suggested that the established roll-to-roll microwave plasma CVD is a promising technique for continuous and large area deposition of graphene films, which is indispensable in order to realize industrial quality transparent conductive films for touch screens and beyond.
93
Acknowledgments The authors would like to thank Dr. K. Tsugawa and Mr. N. Shimada for experimental supports and fruitful discussion. This work was partially supported by the New Energy and Industrial Technology Development Organization (NEDO).
References [1] S. Bae, H. Kim, Y. Lee, X. Xu, J.S. Park, Y. Zheng, J. Balakrishnan, T. Lei, H.R. Kim, Y.I. Song, Y.J. Kim, K.S. Kim, B. Ozyulaz, J.H. Ahn, B.H. Hong, S. Iijima, Nat. Nanotechnol. 5 (2011) 574. [2] H. Yamaguchi, G. Eda, C. Mattevi, H. Kim, M. Chhowlla, ASC Nano 4 (2010) 524. [3] J. Kim, M. Ishihara, Y. Koga, K. Tsugawa, M. Hasegawa, S. Iijima, Appl. Phys. Lett. 98 (2011) 091502. [4] K. Tsugawa, M. Ishihara, J. Kim, H. Hasegawa, Y. Koga, New Diamond Front. Carbon Technol. 16 (2006) 337. [5] A. Kromka, O. Babchenko, T. Izak, K. Hruska, B. Rezek, Vacuum 86 (2012) 776. [6] H. Sumi, S. Ogawa, M. Sato, A. Saikubo, E. Ikenaga, M. Nihei, Y. Takakuwa, Jpn. J. Appl. Phys. 49 (2010) 076201. [7] T. Yamada, J. Kim, M. Ishihara, M. Hasegawa, S. Iijima, Carbon 50 (2012) 2615. [8] L.G. Cancado, K. Takai, T. Enoki, M. Endo, Y.A. Kim, H. Mizusaki, A. Jorio, L.N. Coelho, R. M-Paniago, M.A. Pimenta, Appl. Phys. Lett. 88 (2006) 163106. [9] T. Terasawa, K. Saiki, Carbon 50 (2012) 869. [10] L.M. Malard, M.A. Pimenta, G. Dresselhaus, M.S. Dresselhaus, Phys. Rep. 473 (2009) 51.
Contents lists available at SciVerse ScienceDirect
Thin Solid Films journal homepage: www.elsevier.com/locate/tsf
Large area coating of graphene at low temperature using a roll-to-roll microwave plasma chemical vapor deposition Takatoshi Yamada ⁎, Masatou Ishihara, Masataka Hasegawa Nanotube Research Center, National Institute of Advanced Industrial Science and Technology (AIST), 1-1-1 Higashi, Tsukuba 305-8565, Japan Technology Research Association for Single Wall Carbon Nanotubes (TASC), 1-1-1 Higashi, Tsukuba 305-8565, Japan
a r t i c l e
i n f o
Available online 6 January 2013 Keywords: Graphene Plasma CVD Roll-to-roll deposition Coating Raman spectroscopy
a b s t r a c t Roll-to-roll microwave plasma chemical vapor deposition (CVD) at low temperature has been developed for the fabrication of a continuous and large area deposition of graphene films for the application of transparent conductive films. Deposition of graphene films on copper (Cu) foils, under methane (CH4) and hydrogen (H2) plasma below 380 °C, were confirmed and characterized by Raman spectroscopy. In addition, the film qualities were improved by controlling the CH4/(CH4 + H2) ratio and the process pressure. The graphene film obtained by roll-to-roll microwave plasma CVD process exhibited uniform Raman spectra towards the width direction of the Cu foil (A4 width). Transmittance and haze of the films were confirmed to be uniform and are in sufficient quality for the use of practical touch panel applications. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Transparent conductive films are essential components for many modern electronic and optoelectronic applications such as displays, touch screens, solar cells and light emitting devices. Although the commercial transparent conductive films used for practical applications are made from indium tin oxide (In2O3:Sn, ITO), a substitute material for ITO has been a topic of great interest, due to the limited supply of indium and other rare-metals. In addition, recent advance in organic semiconductor have realized flexible electro-optics devices such as electronic papers and sheet-type solar cells, for which inflexible ITO films are not applicable. Therefore, development of an alternative transparent conductive film has attracted much attention. Graphene, which consists of only carbon (C) atoms, is expected to be one of the most appropriate materials for transparent conductive films, since an ideal mono-layer graphene has a transmittance of 97.7%, an electron mobility of 200,000 cm 2/Vs at room temperature and a thermal conductivity of 5000 W/mK. However, the development of graphene mass production technology has not been established up to now, since it requires either high temperature or long process time. Techniques such as thermal chemical vapor deposition (CVD) on metal catalysis require high temperature (about 1000 °C) and long process time (few tens of minutes) to form graphene films [1]. As for low
⁎ Corresponding author at: Nanotube Research Center, National Institute of Advanced Industrial Science and Technology (AIST), 1-1-1 Higashi, Tsukuba 305-8565, Japan. E-mail address: [email protected] (T. Yamada). 0040-6090/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.tsf.2012.12.102
temperature formation of graphene film, reduction process of graphene oxide was reported [2]. Although this process enables us to form large area graphene on various substrates using spin coating technique, damages and defects during oxidization process still remain in the formed graphene films. Using surface wave plasma CVD, high quality graphene films were obtained at low temperature and few tens of seconds [3]. In addition, touch panel application based on surface wave plasma CVD graphene was demonstrated [3]. From the previous reports, it is expected that the graphene deposition process using surface wave plasma CVD is suitable for an industrial mass production. Roll-to-roll process is another common technique for industrial mass productions of large-area thin film. By combining roll-to-roll process and CVD technique, a continuous deposition process of graphene film is expected. Among various CVD methods, surface wave plasma CVD is selected for combination with roll-to-roll process, because of its possible low-temperature deposition of graphene. In this paper, we describe the development of roll-to-roll microwave plasma (MWP) CVD process towards the industrial mass production of graphene films. In our previous report [3], Ar was used as the carrier gas to keep the plasma stable. However, Ar was considered to damage the graphene and incorporates unintentional impurities from the CVD system. Therefore, effects of flow ratio of CH4 to the total gas (CH4 + H2) and process pressure during CVD process on graphene qualities were examined in this study. A continuous graphene film with 294 mm width is deposited on copper (Cu) foils at low temperature of around 380 °C. High uniformity of transparency (95%) in large area is confirmed. Our data indicated that the roll-to-roll MWPCVD was a promising technique for the graphene mass production towards the transparent conductive films.
90
T. Yamada et al. / Thin Solid Films 532 (2013) 89–93
Fig. 1. The schematic of a roll-to-roll microwave plasma CVD apparatus based on a linear antenna type.
2. Experimental details 2.1. A roll-to-roll system in a linear antenna type microwave plasma CVD for graphene deposition A MWPCVD process based on a linear antenna type system has been one of the potential technique for graphene deposition over large area at low process temperature, because MWP sustained by surface wave mode generates high plasma density (10 11–10 12 cm−3) even at relatively low pressure (b 100 Pa), which is effective to keep the substrate at low temperature. Large area coating of nano-structured diamond films was established using such CVD processes [4,5]. Our linear antenna type MWPCVD system (Fig. 1) had eight coaxial linear antennas covered with quartz tubes, which were used to generate the plasma. The antennas were cooled by air in quartz tubes. The CVD apparatus was equipped with two microwave generators of 2.45 GHz with a maximum power of 20 kW. Source gases (methane: CH4 and hydrogen: H2) were introduced from the top of CVD chamber and evacuated from the bottom of the chamber using a rotary pump [6]. The roll-to-roll system consists of a pair of winder and unwinder. The winder has a motor attached in order to control the flow speed of metal foils in the range from 1 to 500 mm/s. The unwinder is equipped with a brake works to keep the appropriate tension of the metal foils. The sample holder is cooled by water supply. The metal foils are exposed to plasma during traveling of 48 cm length of plasma area. The
Fig. 3. The peak intensity ratio as a function of the CH4/(CH4 + H2) ratio.
numbers of deposited graphene layers can be controlled by foil transfer speed. A rolled 33 μm-thick Cu foils with 294 mm in width, prepared by rolling processes, are used as the substrates in this study. A metal foil of up to 30 m in length and 33 μm in thickness could be installed in our unwinder section [7]. The typical CVD conditions in order to form continuous films were as follows. The microwave power was 12 kW. The temperature of sample stage was less than 380 °C. The flow ratio of CH4 to the total gas (CH4 +H2) [CH4/(CH4 +H2)] and the process pressure was controlled as parameters in order to obtain large domain size. The CH4/(CH4 +H2) ratios were changed from 10 to 70%. In order to keep plasma stably, Ar was used during CVD process in our previous report [7]. However, Ar was not added in this study since Ar was considered to make damage on graphene and to make unintentional impurities from CVD and roll-to-roll system. The pressures in our CVD system were controlled from 30 to 300 Pa. The Cu foil follow speed was 5 mm/s, which let the Cu foil exposed to plasma for 96 s.
2.2. Characterization techniques of graphene films
Fig. 2. Raman spectra of the obtained films, where the CH4/(CH4 + H2) ratios were 10, 30, 50 and 70%.
The obtained graphene films on Cu foils were characterized by Raman spectroscopy excited by a 638 nm diode laser with a beam spot size of 1 μm in diameter. Plane view transmission electron microscopy (TEM) observations were also carried out. Graphene films were mounted on Cu mesh grid. The accelerated voltage was 200 kV. Cross-sectional transmission electron microscopy (XTEM) observations were performed to estimate the length of graphene and the number of layers. In order to prevent unnecessary damage of graphene film prior to the XTEM observation, amorphous carbon was deposited on top of the graphene/Cu substrate. Gallium (Ga) focus ion beam (FIB) etching technique was used for sample preparations. For observations, accelerated voltage was 300 keV and the resolution of TEM used in this study was about 2 nm. In order to evaluate a transmittance, haze and sheet resistance of the film, the substrates (Cu foils) were removed by dipping the sample into a solution of ferric chloride (FeCl3) and the graphene films were transferred onto PET films. The transmittance and the haze at 550 nm were measured by means of visible light spectroscopy. The illuminated area was about 0.7 mm in diameter. Four-probe method was used to measure the sheet resistance. The distance between probes was 1 mm. The measurements were carried out in air at room temperature.
T. Yamada et al. / Thin Solid Films 532 (2013) 89–93
91
Fig. 5. XTEM image of the obtained graphene where the CH4/(CH4 + H2) ratio was 50% and the pressure was 30 Pa.
deposited under the condition where the CH4/(CH4 + H2) ratios were less than 50%, while graphitic carbon film was deposited when the CH4/(CH4 + H2) ratios was 70% [6,8]. As CH4/(CH4 + H2) ratio increases, the number of carbon radicals that reach Cu surface increases. It is speculated that graphitic components formation is dominant rather than graphene formation under the conditions of high carbon radical atmosphere. In order to discuss the quality of the obtained graphene, the D-band to G-band peak intensity ratios (ID/IG) were evaluated, since the ID/IG ratios are related with the average domain size (La) by the following equation [8]: −10 4 −1 La ¼ 2:4 10 λ ðID =IG Þ
ð1Þ
where λ is wave length of laser (638 nm). As increasing the CH4/ (CH4 + H2) ratio, the ID/IG ratio decreases. This indicates that the maximum domain size of the graphene film was obtained by the CVD condition where the CH4/(CH4 + H2) was 50%. The estimated domain size was almost the same as the reported size obtained by plasma CVD [7,9]. Although the lines due to graphite lattice patterns in plane view TEM image as shown in Fig. 4 (a), atomic structures and flake boundaries are not observed, fast Fourier transform (FFT) pattern [Fig. 4 (b)] obtained from Fig. 4 (a) shows hexagonal diffraction pattern and ring pattern, which indicates that the obtained film is graphene. However, the domain length estimated from the XTEM image (as shown in Fig. 5) was in the range of 5 to 15 nm. This is much smaller than that estimated from Raman results. It was reported that the contribution to the peak intensities of D-band from the armchair edge was larger than from the zigzag edges [10]. Therefore, we have to carefully estimate the flake size using Raman results. Although Raman spectroscopy is less powerful to determine the domain length, it enables one to simultaneously evaluate film quality of graphene by using ID/IG ratio, since D-band is attributed to defects. From Raman spectra, we concluded the CVD condition where Fig. 4. (a) Plane view TEM image of the obtained graphene where the CH4/(CH4 + H2) ratio was 50% and the pressure was 30 Pa. And (b) FFT pattern.
3. Results and discussion 3.1. Graphene deposition by linear antenna type microwave plasma CVD In order to investigate the depositing conditions of graphene films on Cu foils using the linear antenna type microwave plasma CVD without the roll-to-roll process, the CH4/(CH4 + H2) ratios and the process pressures were the adjusted parameters. Deposition time was fixed to 90 s for all samples. The films deposited by four kinds of the CH4/(CH4 + H2) ratios at 30 Pa were characterized by Raman spectroscopy (Fig. 2). All Raman spectra show G- and D-bands around 1350 and 1585 cm −1, respectively. Peaks due to 2D-band around 2690 cm −1 are obtained from the films prepared by the CVD conditions where the CH4/(CH4 + H2) ratios were 10, 30 and 50%. Peaks of D′-band are also detected in all spectra. Both of G and 2D-band are attributed to graphene, and D- and D′-bands are originated from defects and/or edges in graphene. By changing the CH4/(CH4 + H2) ratio, the film quality can be modified. The graphene films were
Fig. 6. Raman spectra of the films deposited under pressures at 300 Pa and 30 Pa.
92
T. Yamada et al. / Thin Solid Films 532 (2013) 89–93
Fig. 7. Raman spectra of graphene film prepared by roll-to-roll MWPCVD on Cu foil measured at 4-cm intervals.
the CH4/(CH4 +H2) was 50% is the best condition for our experiment. The 2D-band to G-band peak intensity ratios (I2D/IG) were evaluated to discuss the numbers of layers. From the literature [10], two or three layers were expected after 90 s of CVD process regardless of CH4/(CH4 +H2) concentration. Although it is considered that the deposition rate and size of graphene films increase with CH4/(CH4 +H2) concentration, we concluded CH4/(CH4 +H2) ratio of 50% is the best condition for our experiments, considering the film quality evaluated from Raman spectra. It is speculated that graphitic components formation is dominant rather than graphene formation under the conditions of high carbon radical atmosphere. It must be noted that no signal due to silicon was detected by energy dispersive X-ray spectroscopy (EDX), which will be discussed in elsewhere. The typical Raman spectra obtained from films prepared at 30 and 300 Pa, where the CH4/(CH4 + H2) ratio was 50%, are shown in Fig. 6. It is clear that both spectra show peaks due to G- and D-bands. However, only films obtained at 30 Pa have 2D peak. The Raman data indicate that graphene films were obtained at low process pressure (30 Pa), while graphitic films were deposited at high process pressure (500 Pa). So far, we have not clarified the radical species that are responsible for graphene formation in our plasma CVD. However, it is considered that the first layer, which covers the Cu surface in the beginning, forms graphene but the second layer and up do not form graphene at high process pressure. It is speculated that graphitic component formation is dominant rather than graphene formation under the conditions of high carbon radical atmosphere. From the Raman results as shown in Figs. 3 and 6, both of low CH4/(CH4 + H2) ratio and low process pressure are important to lower the peak intensity of D-band which is attributed to defects or edges.
(1588.2 cm−1), 2D-band (2661.8 cm−1), D-band (1330.8 cm−1) and D′-band (1625.0 cm−1) were observed. Using Eq. (1), the estimated average domain size was about 10 nm. Raman results indicate that the obtained graphene films synthesized by the roll-to-roll microwave plasma CVD consist of small size of domains. It must be noticed that no major differences in Raman spectra compared to our previous reports [3,7], which indicates that the obtained graphene films prepared by roll-to-roll microwave plasma CVD had similar structures with the previous reports [3,7]. The transmittance, the haze and the sheet resistance of the transferred graphene/PET film structures at an interval of 4 cm are shown in Fig. 8. The transmittance includes that of the PET substrates, which is about 91%. The total transmittance is about 89%, the haze is from 0.95 to 1.3%, and the sheet resistance is 0.9–3 × 106 Ω/sq. The obtained transmittance and the sheet resistance are almost the same as those of the reported graphene-based transparent conductive films obtained from the reduction process of graphene oxide [2]. However, the obtained sheet resistance was three or four orders higher than the data of graphene film deposited by thermal CVD. Therefore, the reduction of the sheet resistances is necessary to utilize the practical transparent conductive films in order to use transparent conductive graphene films for practical touch panel applications. The difference in the sheet resistance obtained by MWPCVD and by thermal CVD would be explained by the small crystal flake of the deposited graphene films in the MWPCVD at extremely low temperature compared with that of the thermal CVD, since in the case of the thermal CVD graphene such a high intensity D-band is usually not observed [1]. In addition, the reduction of the sheet resistance by the doping during transfer process was reported [1]. By optimizing the experimental conditions of the CVD process and the transfer process, the reduction of the sheet resistance would be expected. Using surface wave plasma CVD, graphene was deposited on aluminum (Al) foils, which has no catalytic effect to decompose hydrocarbon and hydrogen [3]. Therefore, we expect that graphene can directly be deposited on PET films. Although graphitic carbon films were deposited on PET film by surface wave plasma CVD, we have not yet confirmed graphene formation on PET films. By taking graphene deposition on
3.2. Graphene deposition by roll-to-roll microwave plasma CVD From the obtained results mentioned above, graphene films were deposited by the roll-to-roll microwave plasma CVD under the condition of 50% of the CH4/(CH4 + H2) ratio and 30 Pa of the pressure. Typical Raman spectra of the graphene film at interval of 4 cm in the across-the-width direction of the Cu foil are shown in Fig. 7, which suggests good uniformity of the film. The Raman signals of G-band
Fig. 8. Transmittances, hazes and sheet resistances of roll-to-roll MWPCVD graphene/PET structures measured at 4-cm intervals.
T. Yamada et al. / Thin Solid Films 532 (2013) 89–93
Cu foil into account, it would be expected that graphene deposition on PET film by optimizing the CVD conditions. Up-to-now, the catalytic effect of Cu plays an important role for the deposition of graphene with high rate and low temperature. This is under investigation. 4. Summary We developed a process by combination of the plasma CVD process at low substrate temperature and the roll-to-roll process for the graphene mass production. By selecting the microwave plasma sustained by surface-wave mode, high plasma density (10 11–1012 cm−3) even at relatively low pressure (b 100 Pa) was realized, which is effective to keep the low substrate temperature and to obtain the high deposition rate. It was found from Raman results that the qualities of graphene films were strongly depended on the CH4/(CH4 +H2) ratios and the process pressures. The graphene film obtained by the roll-to-roll microwave plasma CVD process showed uniform Raman spectra towards the width direction of the Cu foil (A4 width). The high uniform transmittance and haze were confirmed, which were enough for the practical touch panel application. By optimizing the conditions of both CVD and transfer process, the improvement of the electrical conductivity will be attained. Our experimental data suggested that the established roll-to-roll microwave plasma CVD is a promising technique for continuous and large area deposition of graphene films, which is indispensable in order to realize industrial quality transparent conductive films for touch screens and beyond.
93
Acknowledgments The authors would like to thank Dr. K. Tsugawa and Mr. N. Shimada for experimental supports and fruitful discussion. This work was partially supported by the New Energy and Industrial Technology Development Organization (NEDO).
References [1] S. Bae, H. Kim, Y. Lee, X. Xu, J.S. Park, Y. Zheng, J. Balakrishnan, T. Lei, H.R. Kim, Y.I. Song, Y.J. Kim, K.S. Kim, B. Ozyulaz, J.H. Ahn, B.H. Hong, S. Iijima, Nat. Nanotechnol. 5 (2011) 574. [2] H. Yamaguchi, G. Eda, C. Mattevi, H. Kim, M. Chhowlla, ASC Nano 4 (2010) 524. [3] J. Kim, M. Ishihara, Y. Koga, K. Tsugawa, M. Hasegawa, S. Iijima, Appl. Phys. Lett. 98 (2011) 091502. [4] K. Tsugawa, M. Ishihara, J. Kim, H. Hasegawa, Y. Koga, New Diamond Front. Carbon Technol. 16 (2006) 337. [5] A. Kromka, O. Babchenko, T. Izak, K. Hruska, B. Rezek, Vacuum 86 (2012) 776. [6] H. Sumi, S. Ogawa, M. Sato, A. Saikubo, E. Ikenaga, M. Nihei, Y. Takakuwa, Jpn. J. Appl. Phys. 49 (2010) 076201. [7] T. Yamada, J. Kim, M. Ishihara, M. Hasegawa, S. Iijima, Carbon 50 (2012) 2615. [8] L.G. Cancado, K. Takai, T. Enoki, M. Endo, Y.A. Kim, H. Mizusaki, A. Jorio, L.N. Coelho, R. M-Paniago, M.A. Pimenta, Appl. Phys. Lett. 88 (2006) 163106. [9] T. Terasawa, K. Saiki, Carbon 50 (2012) 869. [10] L.M. Malard, M.A. Pimenta, G. Dresselhaus, M.S. Dresselhaus, Phys. Rep. 473 (2009) 51.