The effect of substrate size on characteristics of carbon films deposited on glass plate substrates by thermal chemical vapor deposition

Article abstract of DOI:10.1002/pssa.200824496

When carbon films are prepared using thermal CVD, the properties of carbon films are affected by many factors, such as the precursor gas, deposition temperature, and working pressure. It is believed that the substrate size is also an important factor in p

Full text of DOI:10.1002/pssa.200824496

Phys. Status Solidi A 206, No. 7, 1537–1540 (2009) / DOI 10.1002/pssa.200824496
a
p s s
www.pss-a.com
applications and materials science
The effect of substrate size
on characteristics of carbon films
deposited on glass plate substrates
by thermal chemical vapor deposition
Po-Yu Chen1, Sham-Tsong Shiue*, 1, and Hung-Yi Lin2
1
Department of Materials Science and Engineering, National Chung Hsing University, 250 Kuo Kuang Road, Taichung 402, Taiwan
2
Mechanical and Systems Research Laboratories, Industrial Technology Research Institute, Hsinchu 310, Taiwan
Received 12 December 2008, revised 8 April 2009, accepted 22 April 2009
Published online 17 June 2009
PACS 68.37.Hk, 78.30.Na, 81.05.Uw, 81.15.Gh
*
Corresponding author: e-mail stshiue@dragon.nchu.edu.tw, Fax: 886 422857017
When carbon films are prepared using thermal CVD, the mental results show that the deposition thickness decreases
properties of carbon films are affected by many factors, such with increasing silica glass-plate width, indicating that the
as the precursor gas, deposition temperature, and working deposition rate of carbon coatings decreases with increasing
pressure. It is believed that the substrate size is also an impor- deposition area of the glass plate substrate. Meanwhile, the
tant factor in preparing carbon films by thermal CVD. Hence, ordered degree and crystallite size of carbon films increase
this study investigates the effect of substrate size on the prop- with increasing glass-plate width, while the number and size
erties of carbon coatings that are deposited on glass plate sub- of particles on carbon-coating surfaces, surface roughness,
strates in a quartz tube furnace using thermal CVD. Experi- and electrical resistivity of carbon films decrease.
© 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1 Introduction Various properties of carbon films
2 Experimental details The thermal CVD system
were identified, including high water resistance, low used a quartz tube as the reaction furnace, which had a
chemical attack, high hardness, infrared transparency, and length of 400 mm, an internal diameter of 25 mm, and a
wide bandgap. Many studies have employed carbon films wall thickness of 1.5 mm. Four kinds of glass plates with
in commercial applications, such as protective optical ma- an identical length of 120 mm, an identical thickness of
terials, frictional resistance films, biomedical products, and 1 mm, and widths, w, of 3 mm, 6 mm, 12 mm, and 24 mm
others [1, 2]. Carbon films can be deposited using various were adopted as the substrates to prepare the carbon films.
methods, including radio-frequency plasma-enhanced These glass plate substrates were cleaned in ultrasonic
chemical vapor deposition (rf-PECVD) [3], microwave baths of acetone and deionized water, in that order, to im-
plasma CVD [4–6], hot-filament CVD [6], sputtering prove the adhesion of carbon films onto glass plate sub-
deposition [7], filtered cathodic vacuum arc deposition [8], strates. The schematic diagram of the glass plate substrate
ion-beam deposition [9], and thermal CVD [10].
in the quartz tube furnace is shown in Fig. 1. Pure methane
When carbon films are prepared using thermal CVD, (99.999% CH ) was used as the precursor gas with a
4
the properties of carbon films are affected by many factors, flow rate of 40 sccm. The deposition temperature, working
such as the precursor gas, deposition temperature, and pressure, and deposition time were set to 950 1 °C,
working pressure [10–15]. It is believed that the substrate 77 3 kPa, and 30 min, respectively. The temperature was
size is also an important factor in preparing carbon films increased from room temperature to the deposition tem-
by thermal CVD. Hence, this study investigates the effect perature at a rate of 15 °C/min. After the deposition pro-
of substrate sizes on the property of carbon films that are cess, the furnace temperature was quickly reduced to room
deposited on glass plate substrates in a quartz tube furnace. temperature at a rate of 250 °C/min by cooling in air with a
© 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

a
p s s
1538
Po-Yu Chen et al.: The effect of substrate size on characteristics of carbon films
Figure 1 Schematic diagram of the glass plate substrate in a
quartz tube furnace.
fan. The thicknesses of carbon films were obtained by
measuring cross sections of glass plate substrates using a
field emission scanning electron microscope (SEM). The
roughness, particle number, and particle size on carbon
Figure 2 SEM photographs of cross sections of carbon films pre-
pared with different glass-plate widths.
2
film surfaces were obtained over an area of 10 × 10 μm
using an atomic force microscope (AFM) in tapping mode.
Each datum was obtained as an average value from three
positions on the surface of a single specimen. The micro-
structure of the carbon films was investigated using a Ra-
man scattering spectrometer. Raman spectra (RS) were
measured in the back scattering geometry with the 633 nm
line of a He–Ne laser at room temperature in the spectral
film surface, and the particle size decreases with increasing
glass-plate width. In other words, the particle size in-
creases with increasing carbon-film thickness. Moreover,
SEM photographs illustrate that carbon films exhibit lami-
nar structure. The pyrolytic carbon films that are prepared
by thermal CVD usually have the laminar structure [16].
The roughness, particle number, and particle size on car-
bon film surfaces can be obtained by AFM measurement.
Table 1 reveals that the root-mean-square surface rough-
–1
range of 800–1800 cm . The electrical property of the car-
bon films was measured using the four-point probe (FPP)
method at room temperature. A direct current of 1 mA was
applied to the carbon films, and the sheet resistance was
obtained from the relationship between voltage and current.
Each datum was obtained as an average value from five
positions on the surface of a single specimen.
ness (R ) of carbon films are 15.4 nm, 8.4 nm, 4.8 nm,
rms
and 2.5 nm for glass-plate widths of 3 mm, 6 mm, 12 mm,
and 24 mm, respectively, indicating that the outer surfaces
of carbon films become smooth as the glass-plate width in-
creases. Table 1 also shows that the particle number and
average particle size on the carbon film surface decrease
with increasing glass-plate width. This result is consistent
with that shown in SEM photographs of Fig. 2.
3 Results and discussion Figure 2 shows SEM pho-
tographs of cross sections of carbon films that are prepared
with different glass-plate widths. The thicknesses of car-
Using the thickness of carbon films obtained from
bon films are obtained by measuring cross sections of glass SEM and the deposition time, the deposition rate r of the
a
carbon films can be calculated. Table 1 reveals that the
deposition rates are 12.8 nm/min, 12.2 nm/min, 11.6 nm/
min, and 10.8 nm/min for glass-plate widths of 3 mm,
6 mm, 12 mm, and 24 mm, respectively, indicating that the
deposition rate of carbon films decreases with increasing
glass-plate width. As the glass plate size is changed, the
plate substrates. Table 1 shows that the thicknesses of car-
bon films are 383 nm, 367 nm, 347 nm, and 325 nm for
glass-plate widths of 3 mm, 6 mm, 12 mm, and 24 mm, re-
spectively. These results indicate that the thickness of car-
bon films decreases with increasing glass-plate width. Fig-
ure 2 also reveals that many particles appear on the carbon
Table 1 Film thickness, surface roughness, particle number, average particle size, deposition rate, and electrical resistivity of carbon
films prepared with different glass-plate widths.
glass-plate width (mm)
film thickness (nm)
3
6
12
24
383
6
367
5
347
3
325
2.5
3
R
(nm)
15.4
3
8.4 1.5
0.767 0.026
230
4.8 1.2
0.428 0.012
195
1
rms
2
particle number per μm
0.837 0.035
258
0.179 0.012
180
average particle diameter (nm)
deposition rate (nm/min)
12.8 0.2
21.8 1.5
12.2 0.2
11.6 0.1
10.8 0.1
electrical resistivity (Ω μm)
19.5
1
17.8
1
16.3
1
© 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.pss-a.com

Original
Paper
Phys. Status Solidi A 206, No. 7 (2009)
1539
Figure 3 Raman spectra of carbon films prepared with different
glass-plate widths.
average flow velocity v of the precursor gas in the depos-
av
ited zone and the deposition area of the glass plate sub-
strate are also altered. However, because the cross section
of the glass plate is much smaller than that of the quartz
Figure 4 RS data and crystallite size of carbon films prepared
with different glass-plate widths.
tube furnace, the change of the average flow velocity v of
av
the precursor gas in the deposited zone would be very
small. Hence, only the effect of the deposition area of the
glass plate on the deposition rate r is considered here. The
a
2
breathing mode of sixfold sp rings and identified as the
dependence of r on w can be expressed by a power func-
a
n
disorder-activated band. This mode is forbidden in perfect
graphite and only becomes active in the presence of disor-
der. The G-band is associated with the bond-stretching vi-
tion as: r = k (w/w ) , where w (= 3 mm) is the reference
a 1 0 0
glass-plate width, and k and n are constants to be deter-
1
mined. k and n can be obtained by curve fitting, and the
1
2
brations of all sp sites [17]. Table 2 shows the D peak po-
result shows that k and n are 12.8 nm/min and –0.0739,
1
sition (ω ), G-peak position (ω ), full width at half-
D G
respectively. Because the deposition rate r of carbon films
a
maximum of the D-band (FWHM ) and G-band (FWHM ),
D G
decreases with increasing deposition area of the substrate,
the value of n is negative. In the CVD process, the carbon
species in the CVD gas are nucleated on the glass plate sub-
strate at the beginning of the film growth. If the flow rate is
kept constant, the nucleation stage at a larger glass plate
area would require more time to complete, thus, leading to a
reduced film thickness. Therefore, the deposition rate of
carbon films decreases with increasing glass-plate width.
Figure 3 shows the RS of carbon films that are pre-
pared with different glass-plate widths. The RS can be de-
composed into two absorption peaks, a D-band at about
and the ratio of the integrated intensity of the D-band to
that of the G-band (I /I ) for carbon films that were pre-
D
G
pared with different glass-plate widths. The above RS fit-
ting results are also plotted in Fig. 4. This figure reveals
that ω , ω , FWHM , and I /I values decrease with in-
D
G
D
D
G
creasing glass-plate width, while the FWHM value in-
G
creases, indicating that the structural ordered degree of
carbon films increases with increasing glass-plate width
[15, 17–19]. Since the number and size of particles on the
carbon-coating surface decrease with increasing carbon-
film thickness, the ordered degree of carbon films is thus
increased with increasing the glass-plate width. The ratio
–1
–1
1330 cm and a G-band at about 1580 cm , using the
Gaussian function. The D-band is associated with the
Table 2 RS fitting data of carbon films prepared with different glass-plate widths.
glass-plate width (mm)
3
6
12
24
–1
ω (cm )
D
1328.1 1.0
1587.6 0.8
299.4 0.9
106.5 0.6
3.81 0.1
1327.6 0.5
1586 0.2
298.7 0.6
106.9 1.0
3.76 0.03
10.3 0.1
1327.2 1.4
1584.7 1.3
292.8 0.8
109 0.3
1325.1 0.3
1582.8 0.5
289.7 1.1
109.4 0.6
3.43 0.03
11.2 0.1
–1
ω (cm )
G
–1
FWHM (cm )
D
–1
FWHM (cm )
G
I /I
D
3.59 0.04
10.7 0.1
G
L (nm)
a
10.1 0.3
www.pss-a.com
© 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

a
p s s
1540
Po-Yu Chen et al.: The effect of substrate size on characteristics of carbon films
I /I is commonly used to approximately estimate the crys-
D G
References
tallite size of nanographite, and an expression that gives
the crystallite size (L ) from the integrated intensity ratio
[1] A. Grill, Diam. Relat. Mater. 8, 428 (1999).
[2] J. Robertson, Mater. Sci. Eng. R 37, 129 (2002).
[3] S. T. Shiue, H. H. Hsiao, T. Y. Shen, H. C. Lin, and
K. M. Lin, Thin Solid Films 483, 140 (2005).
[4] Y. Liu, C. Liu, Y. Chen, Y. Tzeng, P. Tso, and I. Lin, Diam.
Relat. Mater. 13, 671 (2004).
a
–10
4
–1
I /I is given by [20]: L (nm) = (2.4 × 10 ) λ (I /I ) ,
D
G
a
l
D
G
where λ (in nm units) is the laser line wavelength used in
l
the Raman experiment. This expression indicates that the
crystallite size of nanographite increases with decreasing
I /I value. Hence, Table 2 and Fig. 4 show that I /I de-
[5] O. Brinza, J. Achard, F. Silva, X. Bonnin, P. Barroy, K. De
Corte, and A. Gicquel, Phys. Status Solidi A 205, 2114
(2008).
D
G
D
G
creases from 3.81 to 3.43 as the glass-plate width increases
from 3 mm to 24 mm, while L increases from 10.1 nm to
a
[6] H. Calvo del Castillo, J. L. Ruvalcaba, E. Belmont, T. Cal-
derón, R. Meléndrez, and M. Barboza-Flores, Phys. Status
Solidi A 205, 2221 (2008).
11.2 nm. Notably, the CH bonds in the region around
–1
3000 cm are not observed by Raman scattering and/or in-
frared spectroscopy.
[7] G. Fanchini, A. Tagliaferro, B. Popescu, and E. A. Davis,
J. Non-Cryst. Solids 299–302, 846 (2002).
The sheet resistance of carbon coatings can be obtained
by FPP measurements. The electrical resistivity is the
product of the sheet resistance and thickness [21], which
only depends on the structure of carbon films. Thus, the
electrical resistivity is adopted to evaluate the electrical
property of carbon films. Table 1 shows that the electrical
resistivities of carbon films are 21.8 Ω μm, 19.5 Ω μm,
17.8 Ω μm, and 16.3 Ω μm for glass-plate widths of 3 mm,
6 mm, 12 mm, and 24 mm, respectively, indicating that the
electrical resistivity of carbon films decreases as the glass-
[8] N. D. Baydoğan, Mater. Sci. Eng. B 107, 70 (2004).
[9] M. J. Paterson, Diam. Relat. Mater. 7, 908 (1998).
[10] S. S. Chen, S. T. Shiue, Y. H. Wu, and K. J. Cheng, Surf.
Coat. Technol. 202, 798 (2007).
[11] J. Z. Hu and K. J. Huttinger, Carbon 41, 1501 (2003).
[12] H. Nakayama, K. Takatsuji, S. Moriwaki, K. Murakami,
K. Mizoguchi, and M. Nakayama, Thin Solid Films 430,
309 (2003).
[13] C. A. Taylor, M. F. Wayne, and W. K. S. Chiu, Surf. Coat.
Technol. 182, 131 (2004).
plate width increases. The RS results indicate that L in-
a
[14] W. N. Li, Y. S. Ding, S. L. Suib, J. F. DiCarlo, and
F. S. Galasso, Surf. Coat. Technol. 190, 366 (2005).
[15] Y. S. Ding, W. N. Li, S. Iaconetti, X. F. Shen, J. DiCarlo,
F. S. Galasso, and S. L. Suib, Surf. Coat. Technol. 200,
3041 (2006).
creases as the glass-plate width increases. The increase of
L results in a decrease of the grain boundaries, and thus
a
causes a decrease of the electrical resistivity [22, 23]. Be-
sides, the decrease of the electrical resistivity also results
from the decrease of particles on the carbon film surface.
The laminar structure of carbon films is disturbed by the
particles, so the electrical resistivity of carbon films should
decrease if the number and size of particles on carbon film
surfaces decrease.
[16] X. Bourrat, J. Lavenac, F. Langlais, and R. Naslain, Carbon
39, 2376 (2001).
[17] A. C. Ferrari and J. Robertson, Phys. Rev. B 61, 14095
(2000).
[18] E. Tomasella, C. Meunier, and S. Mikhailov, Surf. Coat.
Technol. 141, 286 (2001).
4 Conclusions Carbon films are deposited on glass
plate substrates with different substrate sizes using thermal
CVD. The thickness of carbon films decreases with in-
creasing glass-plate width, indicating that the deposition
rate of carbon coatings decreases on increasing the deposi-
tion area of the glass plate. The number and size of parti-
cles on carbon film surfaces decrease with increasing
glass-plate width. Meanwhile, the structural ordered degree
and crystallite size of carbon films increase with increasing
glass-plate widths, while the surface roughness and electri-
cal resistivity of carbon films decrease.
[19] M. Lejeune, O. Durand-Drouhin, J. Henocque, R. Bouzerar,
A. Zeinert, and M. Benlahsen, Thin Solid Films 389, 233
(2001).
[20] L. G. Cancado, K. Takai, T. Enoki, M. Endo, Y. A. Kim,
H. Mizusaki, A. Jorio, L. N. Coelho, R. Magalhães-Pani-
ago, and M. A. Pimenta, Appl. Phys. Lett. 88, 163106
(2006).
[21] M. Iwaki, Y. Suzuki, A. Nakao, and H. Watanabe, Nucl.
Instrum. Methods Phys. Res. B, Beam Interact. Mater. At.
127/128, 208 (1997).
[22] T. H. Fang and W. J. Chang, Appl. Surf. Sci. 220, 175
(2003).
[23] J. H. Son, M. Y. Park, and S. W. Rhee, Thin Solid Films
335, 229 (1998).
Acknowledgements This work was supported by the Na-
tional Science Council, Taiwan, under Grant No. NSC 97-2221-
E-005-027-MY2, and was also supported in part by the Ministry
of Education, Taiwan, under the ATU plan.
© 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.pss-a.com