Graphitization of ultrathin amorphous carbon films on

JOURNAL OF APPLIED PHYSICS

VOLUME 88, NUMBER 1

1 JULY 2000

Graphitization of ultrathin amorphous carbon films on Si„001… by Ar¿ ion irradiation at ambient temperature Sang Sub Kima) Department of Materials Science and Metallurgical Engineering, Research and Development Center for Automobile Parts and Materials, Sunchon National University, Sunchon 540-742, Korea

Shunichi Hishita National Institute for Research in Inorganic Materials, Namiki 1-1, Tsukuba, Ibaraki 305-0044, Japan

Tae Sik Cho and Jung Ho Jeb) Department of Materials Science and Engineering, Pohang University of Science and Technology, Pohang 790-784, Korea

共Received 29 December 1999; accepted for publication 3 April 2000兲 We report the graphitization of ultrathin 共8 nm兲 amorphous carbon films on Si共001兲 by 2 MeV Ar⫹ ion irradiation at ambient temperature. The resulting graphite films show the 具001典 preferred orientation in the film normal direction, but random distribution in the film plane direction. The smooth surface and interface suggest that the activation energy for the graphitization is supplied mostly by the electronic stopping process while Ar⫹ ions travel through the amorphous carbon film. © 2000 American Institute of Physics. 关S0021-8979共00兲06513-0兴

共XPS兲, electron beam evaporators, etc. 共001兲-oriented Si single-crystal wafers were used as substrates. After a conventional chemical treatment, the substrates were loaded into the chamber, then subsequently heated at 1473 K under a vacuum pressure of below 5⫻10⫺7 Pa for 1 min, which resulted in typical (2⫻1) reconstruction on the surface, as observed by RHEED, and no trace of contamination, as confirmed by XPS. After obtaining such ideal Si共001兲 surfaces, carbon thin films with a thickness of approximately 8 nm were evaporated onto the substrates at ambient temperature. The film thickness was determined from the x-ray reflectivity curves. The ultrathin carbon films were then irradiated by 2 MeV Ar⫹ ions at ambient temperature with an ion current density of 10 mA/m2 and an ion dose level of 2.8 ⫻1020 ions/m2 using an ion accelerator. The film temperature was monitored by a thermocouple attached to the surface of the film. During the ion irradiation, the temperature rise was only as low as around 50 K. Synchrotron x-ray scattering experiments were carried out at Beamline 5C2 at Pohang Light Source 共PLS兲 at the Pohang Accelerator Laboratory, Korea. The incident x rays were focused in the vertical direction by a bending mirror. A double bounce Si 共111兲 monochromator was used to monochromatize the x rays to 1.771 Å and to focus the beam in the horizontal direction. The momentum transfer resolution was controlled by two pairs of slits in front of the detector, and was set at 0.001 Å⫺1 in this experiment. To achieve the exact scattering geometry, we employed a four-circle x-ray diffractometer that enabled us to obtain an arbitrary momentum transfer q in three-dimensional reciprocal space.

I. INTRODUCTION

Carbon has a wide range of properties depending on its phase. Carbon exists in both crystalline and amorphous phases. Crystalline carbon includes graphite, diamond, and a family of fullerenes.1 Each carbon phase has its own particular characteristics. Graphite’s useful properties include high electrical and thermal conductivity as well as high mechanical strength, increasing with temperature.2 Therefore, there has been interest in preparing its thin-film form and modifying its properties by doping with some elements.3,4 Ion beam irradiation is a well-developed technique for modification of surfaces and near-surface regions of solids. This technique has been also applied to induce recrystallization or phase transformation of amorphous regions.5,6 Because under ambient conditions the stable phase of carbon is graphite, amorphous carbon or diamond may be transformed into graphite if a proper energy is applied. From both practical and theoretical viewpoints, it seems interesting to study whether a metastable phase such as amorphous carbon can be transformed into a stable graphite phase only by ion beam irradiation without applying any other thermal energy. In this work, the ion-beam-induced graphitization of ultrathin amorphous carbon films at ambient temperature is reported. The crystallized graphite phase is observed by reflection high-energy electron diffraction 共RHEED兲 and synchrotron x-ray scattering. II. EXPERIMENTAL DETAILS

Carbon deposition and ion irradiation in this work were carried out in an ultra high vacuum 共UHV兲 chamber system equipped with RHEED, x-ray photoelectron spectroscopy

III. RESULTS AND DISCUSSION

The as-deposited carbon films were analyzed by RHEED to figure out what structure they have. As expected, no visible intensity distributions were found in the RHEED pat-

a兲

Electronic mail: [email protected] Electronic mail: [email protected]

b兲

0021-8979/2000/88(1)/55/4/$17.00

55

© 2000 American Institute of Physics

Downloaded 23 Dec 2004 to 168.131.142.104. Redistribution subject to AIP license or copyright, see http://jap.aip.org/jap/copyright.jsp

56

J. Appl. Phys., Vol. 88, No. 1, 1 July 2000

Kim et al.

FIG. 1. RHEED pattern of an ultrathin amorphous carbon film irradiated by Ar⫹ ions. The spots along the specular rod and regularly spaced diffuse lines reveal that the amorphous state of the film is changed into a turbostratic graphite structure by ion irradiation at ambient temperature.

terns, demonstrating that the as-deposited films were amorphous. The amorphous state was also confirmed using XPS in our previous work.7 After Ar⫹ ion irradiation to such amorphous carbon films at ambient temperature, however, there was a sharp difference in the RHEED patterns. Figure 1 shows a typical RHEED pattern obtained from an ionirradiated carbon film. The spot-like specular rod and equally spaced diffuse lines suggest that the amorphous state of the ultrathin carbon film was transformed into a ‘‘turbostratic’’ crystallized structure. It is worth mentioning that graphitization occurs at ambient temperature only by the irradiation effect, without any additional thermal energy. The terminology of the turbostratic structure has been used to describe a graphite structure in which planar sheets of carbon atoms are stacked regularly with a constant distance, while there is no relationship between adjacent lattices in the planar direction. As illustrated in Fig. 1, from the spacing, c ⬘ between each spot along the specular rod, the lattice parameter, c of the graphite film in the film’s normal direction can be estimated as about 0.67 nm using c ⬘ ⫽k2/c. Meanwhile, from the spacing, a ⬘ between the diffuse lines, the in-plane lattice parameter, a, can be roughly estimated as about 0.23 nm using a ⬘ ⫽k)/a. k was determined using (2⫻1) pattern of Si共001兲 as a standard. Note that the lattice parameters estimated from the spacings have some errors because, as can be seen, the spots and lines are broad and diffuse, thereby causing difficulty in defining exact spacings. Nevertheless, these values are quite similar to those of a bulk graphite 共a ⫽0.2463 nm, c⫽0.6714 nm兲,8 which again supports the fact that ion irradiation leads to graphitization of the amorphous carbon film even at ambient temperature without any additional thermal energy. The RHEED technique reveals a surface structure limited only in a few atomic layers from the top surface. We thus employed synchrotron x-ray scattering to investigate the microstructure of the whole film. We first carried out the conventional powder diffraction measurement to obtain the crystal structure and the out-of-plane orientation of the film, by varying the momentum transfer qz along the substrate normal direction. Figure 2 illustrates a typical powder dif-

FIG. 2. Powder diffraction pattern of an ultrathin graphite film on Si共001兲 obtained by Ar⫹ ion irradiation at ambient temperature. The rocking curve obtained at the graphite 共002兲 Bragg reflection is shown in the inset.

fraction pattern of the ion-irradiated carbon films. It is worth noting that there exists one definite Bragg reflection at qz ⫽1.809 Å ⫺1 , which corresponds to the graphite 共002兲 reflection, besides the Si共004兲 reflection. This result clearly demonstrates that the as-deposited amorphous carbon phase is successfully transformed into the crystalline graphite phase by the Ar⫹ ion irradiation treatment at ambient temperature. We also note that the full width at half maximum 共FWHM兲 of the graphite 共002兲 reflection is relatively broad as ⌬qz⫽0.113 Å ⫺1 as shown in Fig. 2. From the FWHM of the graphite 共002兲 reflection, we can estimate the crystal domain size of the graphite grains using Sherrer’s formula.9 The crystal domain size of the graphite grains is 6 nm in the film normal direction, a little less than the film thickness 共8 nm兲. Since the grain size is usually greater than the crystal domain size, probably due to defects within grains, we conjecture that the crystalline graphite grains are almost single grained in the film normal direction. The mosaic distribution of the crystalline axis, which represents the crystalline quality of a thin film, is usually studied by measuring the rocking curve of the Bragg reflection. The mosaic distribution of the graphite 共002兲 grains is 4.02° 共FWHM兲, as illustrated in the inset of Fig. 2. This result clearly indicates that the crystalline graphite film has the 共001兲 preferred orientation on the Si共001兲 substrate in the film normal direction. The graphite 共002兲 grains, however, are randomly distributed in the film plane direction 共data not shown兲. The surface and interface roughnesses of the graphite/ Si共001兲 film provide an important basis by which to elucidate graphitization of the amorphous carbon films during Ar⫹ ion irradiation. To study the surfacial and interfacial behavior, we performed an x-ray reflectivity measurement that is one of the powerful methods by which to study morphological properties of buried interfaces.10 The specular x-ray reflectivity represents the interference pattern of the x rays re-


Kim et al.


FIG. 3. 共a兲 X-ray reflectivity measurements of an ultrathin graphite film on Si共001兲 obtained by Ar⫹ ion irradiation at ambient temperature. The open and closed circles represent the total reflectivity and its diffuse component, respectively. 共b兲 Specular component of the reflectivity. A fitting result is displayed by the solid line.

flected from the surface and the interface, which is described by11 S spec共 q兲 ⫽

A q z2

2 2

2 2

关 ␳ 21 e ⫺q z ␴ 1 ⫹ 共 ␳ 2 ⫺ ␳ 1 兲 2 e ⫺q 1 ␴ 2 2

2

2

⫹2 ␳ 1 共 ␳ 2 ⫺ ␳ 1 兲 cos共 q z d 兲 e ⫺q z 共 ␴ 1 ⫹ ␴ 2 兲 /2兴 ␦ 共 q储兲 , where q z (q储 ) is the x-ray momentum transfer along the surface normal 共in-plane兲 direction, ␳ 1 ( ␳ 2 ) the electron density of graphite 共Si兲, ␴ 1 ( ␴ 2 ) the root-mean-squared surface 共interface兲 roughness, and d the average film thickness. Figure 3共a兲 shows the x-ray reflectivity data. The total reflectivity measured on the specular rod consists of the signals from the specular component as well as those from the diffuse component. The diffuse scattering was measured 0.1° away from the specular rod. It is of note that the intensity oscillations, having originated from the interference of the x rays reflected from the surface and those reflected from the graphite/Si interface, are observed not only in the total reflectivity, but also in the diffuse reflectivity with the same period. This indicates that the surface and interface between the film and the substrate correlate well with each other and are smooth enough to induce such well-defined oscillations. The specular component of the x-ray reflectivity, obtained by subtracting the diffuse component from the total component, is shown in Fig. 3共b兲. The thickness of the graphite film, determined by the period of the intensity oscillations ⌬q in the reflectivity curve by 2 ␲ /⌬q, is 8 nm. The oscillation amplitude and the overall x-ray intensity are determined by the roughnesses of the surface and of the interface.12 To quantify the roughnesses of the surface and of

57

the interface, we analyzed the reflectivity data by using the distorted wave Born approximation 共DWBA兲.12 The solid line in Fig. 3共b兲 is the result of the fitting using the DWBA. This result shows that the roughnesses of the surface and the graphite/Si interface are as small as 0.83 and 1 nm, respectively. This clearly confirms that both the surface and the interface are very smooth. The smooth roughnesses of the surface and the interface and their good correlation imply that the Ar⫹ ion irradiation leading to graphitization induces no significant atomic mixing at the surface and interface regions. Some activation energy should be supplied for the phase transformation of the amorphous carbon into the crystalline graphite structure. Normally ion irradiation provides energy to irradiated solids through nuclear collision or an inelastic electronic stopping process. It is very likely that the energy transfer through nuclear collision leads to substantial atomic mixing at surface and interface regions. In this case quite a rough surface and/or diffuse interface would be expected. In fact, both the surface and the interface of an Ar⫹ ionirradiated carbon film are very smooth. This result suggests that the activation energy required for phase transformation of the amorphous carbon into the graphite phase is supplied predominantly by the inelastic electronic stopping process. According to the Monte Carlo simulation code TRIM, it is reported that the electronic stopping or inelastic energy loss process is a dominant energy transfer process when Ar⫹ ions with an energy of 2 MeV irradiate a carbon solid.13 IV. CONCLUSIONS

In summary, we found ultrathin 共8 nm兲 amorphous carbon films to be transformed into crystallized graphite films by 2 MeV Ar⫹ ion irradiation at ambient temperature using RHEED and synchrotron x-ray scattering. The transformed graphite films showed the 共001兲 preferred orientation without any in-plane alignment. The reflectivity data revealed that the surfaces and interfaces between the film and substrate were very smooth, suggesting that the activation energy for phase transformation is transferred to the amorphous carbon films mainly through the electronic stopping loss process, as was expected using the TRIM simulation. ACKNOWLEDGMENTS

One of the authors 共S.S.K.兲 acknowledges financial support for his stay in Japan by the Science and Technology Agency, Japan. The authors acknowledge partial financial support from the Research Center for Automobile’s Parts and Materials at Sunchon National University, Korea. The PLS is supported by Korean Ministry of Science and Technology. B. Bhushan, Diamond Relat. Mater. 8, 1985 共1999兲. S. Glenis, A. J. Nelson, and M. M. Labes, J. Appl. Phys. 86, 4464 共1999兲. 3 D. I. Jones and A. D. Stewart, Philos. Mag. B 46, 423 共1982兲. 4 J. H. Kaufman, S. Merit, and D. D. Saperstein, Phys. Rev. B 39, 13053 共1989兲. 5 V. Heera, J. Stoemenos, R. Kogler, and W. Skorupa, J. Appl. Phys. 77, 2999 共1995兲. 1 2


58 6

Kim et al.


R. Nowak, F. Yoshida, J. Morgiel, and B. Major, J. Appl. Phys. 85, 841 共1998兲. 7 S. Hishita, K. Oyoshi, S. Suehara, and T. Aizawa, Nucl. Instrum. Methods 148, 594 共1999兲. 8 JCPDS 共Joint Committee on Powder Diffraction Standards兲 Card No. 23– 64. 9 B. E. Warren, X-ray Diffraction 共Addison–Wesley, Reading, MA, 1969兲, Chap. 13.

10

D. Y. Noh, Y. Hwu, H. K. Kim, and M. Hong, Phys. Rev. B 51, 4441 共1995兲. 11 S. K. Shinha, M. K. Sanyal, S. K. Satija, C. F. Majkzak, D. A. Neumann, H. Homma, S. Szpala, A. Gibaud, and H. Morkoc¸, Physica B 198, 72 共1994兲. 12 S. K. Shinha, E. B. Sirota, S. Garoff, and H. B. Stanley, Phys. Rev. B 38, 2297 共1988兲. 13 J. F. Ziegler and J. P. Biersack, TRIM Ver. 96 共1996兲.
