- Email: [email protected]
Influence of the methane concentration on HF–CVD diamond under atmospheric pressure
Vacuum 63 (2001) 449–454
Influence of the methane concentration on HF–CVD diamond under atmospheric pressure Kenji K. Hirakuria,*, Toshihiro Kobayashia, Eri Nakamuraa, Nobuki Mutsukurab, Gernot Friedbacherc, Yoshio Machib a
Department of Electronic and Computer Engineering, Faculty of Science and Engineering, Tokyo Denki University, Ishizaka, Hatoyama, Saitama 350-0394, Japan b Department of Electronic Engineering, Faculty of Engineering, Tokyo Denki University, 2-2 Chiyoda-ku, Kandanishiki-cho, Tokyo 101-8457, Japan c Institute of Analytical Chemistry, Vienna University of Technology, Getreidemarkt 9/151, A-1060 Wien, Austria Received 13 February 2001; received in revised form 9 April 2001
Abstract The most common approach for chemical vapor deposition (CVD) of diamond is the utilization of hydrocarbon gases highly diluted in hydrogen at low pressure (e.g. several thousands of Pascals (Pa)). The quality and growth rate of diamond strongly depends on the methane gas concentration, especially at high pressure, because the generation of atomic hydrogen sharply decreases with increasing pressure. In order to increase the growth rate, we have carried out CVD diamond growth under atmospheric pressure. A dramatic increase of the growth rate could be achieved when using the hot-filament (HF)–CVD technique at atmospheric pressure. Such an increase could already be observed in a previous experiment, however, under varying pressure and at a constant methane concentration of 0.5%. Furthermore, the crystalline quality of the diamond grains could be improved by hydrogen etching at atmospheric pressure. In the current study, the methane volume concentration was varied from 0.03% to 2.0% in order to estimate its effect on diamond growth. The relationship between the quality of the deposited diamond and the methane concentration has been investigated by Raman spectroscopy. The amount of activated hydrogen was estimated from the etching rate of non-diamond components. At high atmospheric pressure, high growth rates could be achieved up to a methane concentration of 0.3%. Moreover, the growth rate has also been shown to depend on the residence time of the precursor in the reactor. Finally, Raman analysis revealed an increasing quality of diamond with decreasing methane concentration. r 2001 Elsevier Science Ltd. All rights reserved. Keywords: Diamond growth and characterization; Hot filament; Grain size; Morphology
1. Introduction Chemical vapor deposition (CVD) of diamond films has received considerable attention due to *Corresponding author. Fax: +81-492-96-6413. E-mail address: [email protected] (K.K. Hirakuri).
their attractive film properties such as high thermal conductivity, mechanical hardness, wide band gap, and chemical inertness [1–3]. If rapid growth and improved quality of CVD diamonds can be realized then they are expected to be utilized widely industrially in the fields of electronics, mechanics, and optics.
0042-207X/01/$ - see front matter r 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 2 - 2 0 7 X ( 0 1 ) 0 0 3 6 5 - 7
450
K.K. Hirakuri et al. / Vacuum 63 (2001) 449–454
Up to now most researchers commonly use pressure, flow rate, gas ratios, substrate bias, power and substrate temperature to characterize the growth conditions. As these parameters are not the only crucial deposition parameters a comparison of results achieved with experimental set-ups and commercial systems is normally very much required. Therefore, another standardizeable parameter is required. In our previous work, the residence time was proposed as a well controllable parameter for the CVD diamond growth and the results of the two different systems were in good agreement with the proposed parameter [3]. A significant enhancement in the growth rate of CVD diamond was obtained under atmospheric pressure [3]. Furthermore, the XRD analysis revealed that the quality of diamonds grown at atmospheric pressure was better than that of diamonds produced at low pressure. Takeuchi et al. reported that homoepitaxial growth of diamond with good quality in terms of crystal structure analysis has been achieved with microwave assisted plasma CVD at relatively low methane concentrations [4,5]. Due to ideal crystal structure and high purity, homoepitaxial diamond films can be applied to semiconductor materials. However, the growth rate, depending on the volume of source gas, is in the order of 1 nm min@1 (comparable to molecular beam epitaxy, and MBE) which is too low to obtain useful film thicknesses within a reasonable period of time. In order to overcome this problem, in this work CVD diamond growth under atmospheric pressure has been investigated addressing the effects of methane concentration and hydrogen etching on the growth rate and quality of the diamonds as determined by Raman spectroscopy.
2. Experimental procedures A conventional hot filament CVD system has been used for diamond growth under standard conditions as described previously [3]. The system is equipped with accessories for high pressure protection and thermal safety. The vacuum chamber made of stainless steel contains a molybdenum substrate holder with a thermocou-
ple and a tungsten filament. The vacuum chamber was evacuated by a rotary pump to a background pressure of 0.1 Pa after the substrates were mounted on the substrate holder. The hydrogen gas was introduced into the chamber through a mass flow controller at atmospheric pressure. The filament was heated up to 2370 K by passing an electric current through it and the methane gas was mixed with hydrogen after a stable filament temperature was reached. The substrate temperature was kept at 1120 K when the substrates were located at a distance of 5 mm from the filament. The diamond growth was controlled by the residence time tr (s) of the precursor which is given by tr ¼ k t
PV ; Q
where P is the gas pressure (Pa), Q the flow rate (sccm), V the volume of the deposition chamber (cm3 ), and kt the constant value of 10.5 [3]. In order to investigate the influence of the methane concentration on the quality of the grown diamond, it was varied from 0.03% to 2.0%. In order to study etching effects on the grown diamonds and other deposits, a hot-filament (HF) system was utilized in the same chamber. For this purpose hydrogen gas was introduced into the chamber after the sample was fixed on the sample holder. The pressure of H2 was kept at 10 Pa. The filament and the substrate temperatures were maintained at 2370 and 1120 K, respectively. The residence time of hydrogen gas was varied from 0.5 to 1.5 s. In the case of etching treatment, two residence times were used and the treatment time was kept for 30 min. The surface morphology of the grown diamond was investigated using scanning electron microscopy (SEM: JEOL JSM-5310LV, resolution=4 nm). Micro Raman spectroscopy (JASCO: NSR-2100) was used to identify the quality of the diamond films grown. The growth rate was estimated by the size of the diamond grains and the growth time.
3. Results and discussion Fig. 1 shows the growth rate of diamond versus residence time. At low pressure conditions (5 and
K.K. Hirakuri et al. / Vacuum 63 (2001) 449–454
451
Fig. 1. The growth rate of CVD diamond grains versus residence time for different pressures.
Fig. 2. The growth rate of CVD diamond grains versus methane concentration under atmospheric pressure.
50 kPa), the maximum growth rate was achieved at a residence time of 0.05 s. It is striking to note that under low pressure the dependency follows similar curves, indicating that the diamond growth cannot be characterized by pressure and flow rate alone. Thus, the residence time is suggested to be used as a standardized parameter for CVD diamond growth. With increasing pressure up to atmospheric pressure and above, the growth rate dramatically increases with residence time. The maximum growth rate at 200 kPa was about four times higher than that at low pressure conditions. The reason for this increase is discussed in a previous paper [3]. The effect of the increased growth rate is because of the increased amount of atomic hydrogen and diamond precursors on the substrate surface resulting from a prolonged residence time. Fig. 2 shows the growth rate of diamond versus methane concentration under atmospheric pressure (100 kPa) for an optimized residence time of 1.7 s. It can be seen that the growth rate increases monotonously with concentration up to 0.3%, and then approaches a saturation level. At a low pressure of 3 kPa, the growth rate is too slow to deposit a useful thickness at low methane concentrations (e.g. 30 nm h@1 for a CH4 concentration of 0.05%) [4]. A linear relation between growth rate and methane concentration can be
observed for concentrations lower than 0.25%. This can be explained by the decrease of activated species (such as atomic hydrogen and methyl radical) and the increase of etching since the relative concentration of atomic hydrogen increases with decreasing methane concentration [6]. In order to increase the growth rate for high quality diamond, deposition at low methane concentration under atmospheric pressure has been attempted. Thereby, a high growth rate of 0.18 mm h@1 has been achieved at a methane concentration of 0.03%, which is an increase by a factor of about six compared to low pressure conditions [5]. A reasonable explanation is the competition between the generation and recombination of atomic hydrogen because the absolute generation of atomic hydrogen decreases with the total pressure in the chamber (see Table 1). SEM micrographs of diamond particles deposited on Si substrates under various CH4 concentrations are shown in Fig. 3. The total growth time and the residence time were fixed at 3 h and 1.7 s for all samples, respectively. At low methane concentration single crystals without any facets can be observed (Fig. 3(a)). In contrast, twin structures can be observed on the diamond particles grown at high methane concentrations as shown in Fig. 3(b)–(d). At low methane concentration, the increase in the ratio between
452
K.K. Hirakuri et al. / Vacuum 63 (2001) 449–454
atomic hydrogen and methyl radical enhanced the crystalline property in the diamond grains. In order to compare the quality of the diamonds grown at different conditions, micro Raman spectroscopy was applied. Fig. 4 shows the Raman Table 1 Growth conditions for CVD diamond deposition under atmospheric pressure Parameters
Range
Substrate material Pressure (kPa) Residence time (s) Methane concentration (%) Deposition time (h) Substrate temperature (K)
Si (1 0 0) 5B200 0.02B5 0.03, 0.1, 0.3, 0.5, 1.0, 1.5, 2.0 3 1120
spectra of diamond grains produced at methane concentrations between 0.1% and 1.0%. In each case the grains were identified as diamond due to the sharp peak at 1332 cm@1. Significant peaks at 1350 cm@1 (D band) and 1530 cm@1 (G band) originating from disordered and glassy amorphous carbon around the grains cannot be observed in the Raman spectra. The full width at half maximum (FWHM) of the peaks was used to characterize the quality of the diamond grains. Typical FWHMs of the diamond peak at 1332 cm@1 are listed in Table 2. A significant increase of the FWHM with increasing methane concentration can be observed. In particular, a strong decrease of the FWHM can be found for a methane concentration of 0.5%.
Fig. 3. SEM images showing the morphology of diamond grains grown on a silicon substrate. Methane concentration: (a) 0.1%, (b) 0.3%, (c) 0.5%, and (d) 1.0%. The total deposition time was 3 h in all the cases.
453
K.K. Hirakuri et al. / Vacuum 63 (2001) 449–454
Fig. 4. Raman spectra of CVD diamond grains grown at different methane concentrations. Table 2 FWHMs of Raman peaks at 1333 cm@1 recorded on diamond grains grown at different methane concentrations Methane concentration (%)
FWHM (cm@1)
1.0 0.5 0.3 0.1
11.53 11.42 10.28 7.79
This result is similar to the dependence observed for the growth rate as a function of methane concentration as shown in Fig. 2. At low methane concentrations, the comparative ratio between atomic hydrogen and methyl radical increases with decreasing methane concentration. Thus, non-diamond species can be etched off, which is also supported by the SEM micrographs in Fig. 3. Next, the quality of diamond grains before and after etching was investigated by Raman spectroscopy since non-diamond components can be identified by this technique. All the spectra have been recorded on the same individual diamond grain by focussing the laser beam at the center of the grain (Fig. 5). In the spectrum of the asdeposited sample (at a methane concentration of 0.5%), besides the dominant diamond peak at 1332 cm@1, a weak peak around 1450 cm@1 can be observed, which indicates the existence of amorphous carbon [7,8]. After the etching treatment the amorphous carbon peak is strongly
Fig. 5. Raman spectra of diamond grains before and after etching. Etching I: residence time=0.5 s, etching II: residence time=1.5 s.
Table 3 FWHMs of Raman peaks at 1333 cm@1 recorded on diamond grains before and after etching
Reference After etching I After etching II
Residence time (s)
FWHM (cm@1)
As deposited 0.5 1.5
11.23 6.58 5.76
reduced. Moreover, the FWHMs for the diamond peaks of the as-deposited and the etched samples were determined as well (Table 3). A reduction of the FWHM by the etching treatment from 11.28 to 6.58 cm@1 can be observed. Furthermore, the residence time of the etching gas also influences the FWHM of the etched samples. While an FWHM of 6.58 cm@1 is observed for the sample etched with a residence time of 0.5 s (denoted as etching I in Fig. 5), the respective value for a residence time of 1.5 s (etching II in Fig. 5) is 5.76 cm@1. This can be explained by the increased generation of atomic hydrogen through extension of the residence time, enhancing the removal of non-diamond components [3]. 4. Conclusions Using the HF–CVD technique, the influence of the methane concentration on diamond growth
454
K.K. Hirakuri et al. / Vacuum 63 (2001) 449–454
was investigated under atmospheric pressure. A decreasing growth rate of the diamond could be observed with decreasing methane concentration. A variation of the CH4 concentration from 0.03% to 1.0% leads to an increase in growth rate by a factor of about 10. On the other hand, the quality of the diamond decreases at the same time. However, the significant enhancement of the growth rate at atmospheric pressure allows production of high quality diamond at low CH4 concentration and still with reasonable deposition rates, making the technique useful for industrial applications.
Acknowledgements The authors would like to thank Mr. T. Toya for performing SEM analysis. This work has been
partially supported by the Center for Research of Tokyo Denki University. References [1] Rye RR. J Appl Phys 1994;76:1220–7. [2] Connell LL, Fleming JW, Chu H-N, Vestyck Jr DJ, Jansen E, Butler JE. J Appl Phys 1995;78:3622–30. [3] Kobayashi T, Hirakuri KK, Mutsukura N, Machi Y. Diamond Related Mater 1999;8:1057–60. [4] Takeuchi D, Yamanaka S, Watanabe H, Sawada S, Ichinose H, Okushi H, Kajimura K. Diamond Related Mater 1999;8:1046–9. [5] Takeuchi D, Watanabe H, Yamanaka S, Okushi H, Kajimura K. Phys Stat Sol 1999;174:101–15. [6] Hayashi K, Yamanaka S, Watanabe H, Sekiguchi T, Okushi H, Kajimura K. J Cryst Growth 1998;183:338–46. [7] Hirakuri KK, Yoshii M, Friedbacher G, Grasserbauer M. Diamond Related Mater 1997;6:1031–5. [8] Leeds SM, Davis TJ, May PW, Pickard CDO, Ashfold MNR. Diamond Related Mater 1998;7:233–7.
Influence of the methane concentration on HF–CVD diamond under atmospheric pressure Kenji K. Hirakuria,*, Toshihiro Kobayashia, Eri Nakamuraa, Nobuki Mutsukurab, Gernot Friedbacherc, Yoshio Machib a
Department of Electronic and Computer Engineering, Faculty of Science and Engineering, Tokyo Denki University, Ishizaka, Hatoyama, Saitama 350-0394, Japan b Department of Electronic Engineering, Faculty of Engineering, Tokyo Denki University, 2-2 Chiyoda-ku, Kandanishiki-cho, Tokyo 101-8457, Japan c Institute of Analytical Chemistry, Vienna University of Technology, Getreidemarkt 9/151, A-1060 Wien, Austria Received 13 February 2001; received in revised form 9 April 2001
Abstract The most common approach for chemical vapor deposition (CVD) of diamond is the utilization of hydrocarbon gases highly diluted in hydrogen at low pressure (e.g. several thousands of Pascals (Pa)). The quality and growth rate of diamond strongly depends on the methane gas concentration, especially at high pressure, because the generation of atomic hydrogen sharply decreases with increasing pressure. In order to increase the growth rate, we have carried out CVD diamond growth under atmospheric pressure. A dramatic increase of the growth rate could be achieved when using the hot-filament (HF)–CVD technique at atmospheric pressure. Such an increase could already be observed in a previous experiment, however, under varying pressure and at a constant methane concentration of 0.5%. Furthermore, the crystalline quality of the diamond grains could be improved by hydrogen etching at atmospheric pressure. In the current study, the methane volume concentration was varied from 0.03% to 2.0% in order to estimate its effect on diamond growth. The relationship between the quality of the deposited diamond and the methane concentration has been investigated by Raman spectroscopy. The amount of activated hydrogen was estimated from the etching rate of non-diamond components. At high atmospheric pressure, high growth rates could be achieved up to a methane concentration of 0.3%. Moreover, the growth rate has also been shown to depend on the residence time of the precursor in the reactor. Finally, Raman analysis revealed an increasing quality of diamond with decreasing methane concentration. r 2001 Elsevier Science Ltd. All rights reserved. Keywords: Diamond growth and characterization; Hot filament; Grain size; Morphology
1. Introduction Chemical vapor deposition (CVD) of diamond films has received considerable attention due to *Corresponding author. Fax: +81-492-96-6413. E-mail address: [email protected] (K.K. Hirakuri).
their attractive film properties such as high thermal conductivity, mechanical hardness, wide band gap, and chemical inertness [1–3]. If rapid growth and improved quality of CVD diamonds can be realized then they are expected to be utilized widely industrially in the fields of electronics, mechanics, and optics.
0042-207X/01/$ - see front matter r 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 2 - 2 0 7 X ( 0 1 ) 0 0 3 6 5 - 7
450
K.K. Hirakuri et al. / Vacuum 63 (2001) 449–454
Up to now most researchers commonly use pressure, flow rate, gas ratios, substrate bias, power and substrate temperature to characterize the growth conditions. As these parameters are not the only crucial deposition parameters a comparison of results achieved with experimental set-ups and commercial systems is normally very much required. Therefore, another standardizeable parameter is required. In our previous work, the residence time was proposed as a well controllable parameter for the CVD diamond growth and the results of the two different systems were in good agreement with the proposed parameter [3]. A significant enhancement in the growth rate of CVD diamond was obtained under atmospheric pressure [3]. Furthermore, the XRD analysis revealed that the quality of diamonds grown at atmospheric pressure was better than that of diamonds produced at low pressure. Takeuchi et al. reported that homoepitaxial growth of diamond with good quality in terms of crystal structure analysis has been achieved with microwave assisted plasma CVD at relatively low methane concentrations [4,5]. Due to ideal crystal structure and high purity, homoepitaxial diamond films can be applied to semiconductor materials. However, the growth rate, depending on the volume of source gas, is in the order of 1 nm min@1 (comparable to molecular beam epitaxy, and MBE) which is too low to obtain useful film thicknesses within a reasonable period of time. In order to overcome this problem, in this work CVD diamond growth under atmospheric pressure has been investigated addressing the effects of methane concentration and hydrogen etching on the growth rate and quality of the diamonds as determined by Raman spectroscopy.
2. Experimental procedures A conventional hot filament CVD system has been used for diamond growth under standard conditions as described previously [3]. The system is equipped with accessories for high pressure protection and thermal safety. The vacuum chamber made of stainless steel contains a molybdenum substrate holder with a thermocou-
ple and a tungsten filament. The vacuum chamber was evacuated by a rotary pump to a background pressure of 0.1 Pa after the substrates were mounted on the substrate holder. The hydrogen gas was introduced into the chamber through a mass flow controller at atmospheric pressure. The filament was heated up to 2370 K by passing an electric current through it and the methane gas was mixed with hydrogen after a stable filament temperature was reached. The substrate temperature was kept at 1120 K when the substrates were located at a distance of 5 mm from the filament. The diamond growth was controlled by the residence time tr (s) of the precursor which is given by tr ¼ k t
PV ; Q
where P is the gas pressure (Pa), Q the flow rate (sccm), V the volume of the deposition chamber (cm3 ), and kt the constant value of 10.5 [3]. In order to investigate the influence of the methane concentration on the quality of the grown diamond, it was varied from 0.03% to 2.0%. In order to study etching effects on the grown diamonds and other deposits, a hot-filament (HF) system was utilized in the same chamber. For this purpose hydrogen gas was introduced into the chamber after the sample was fixed on the sample holder. The pressure of H2 was kept at 10 Pa. The filament and the substrate temperatures were maintained at 2370 and 1120 K, respectively. The residence time of hydrogen gas was varied from 0.5 to 1.5 s. In the case of etching treatment, two residence times were used and the treatment time was kept for 30 min. The surface morphology of the grown diamond was investigated using scanning electron microscopy (SEM: JEOL JSM-5310LV, resolution=4 nm). Micro Raman spectroscopy (JASCO: NSR-2100) was used to identify the quality of the diamond films grown. The growth rate was estimated by the size of the diamond grains and the growth time.
3. Results and discussion Fig. 1 shows the growth rate of diamond versus residence time. At low pressure conditions (5 and
K.K. Hirakuri et al. / Vacuum 63 (2001) 449–454
451
Fig. 1. The growth rate of CVD diamond grains versus residence time for different pressures.
Fig. 2. The growth rate of CVD diamond grains versus methane concentration under atmospheric pressure.
50 kPa), the maximum growth rate was achieved at a residence time of 0.05 s. It is striking to note that under low pressure the dependency follows similar curves, indicating that the diamond growth cannot be characterized by pressure and flow rate alone. Thus, the residence time is suggested to be used as a standardized parameter for CVD diamond growth. With increasing pressure up to atmospheric pressure and above, the growth rate dramatically increases with residence time. The maximum growth rate at 200 kPa was about four times higher than that at low pressure conditions. The reason for this increase is discussed in a previous paper [3]. The effect of the increased growth rate is because of the increased amount of atomic hydrogen and diamond precursors on the substrate surface resulting from a prolonged residence time. Fig. 2 shows the growth rate of diamond versus methane concentration under atmospheric pressure (100 kPa) for an optimized residence time of 1.7 s. It can be seen that the growth rate increases monotonously with concentration up to 0.3%, and then approaches a saturation level. At a low pressure of 3 kPa, the growth rate is too slow to deposit a useful thickness at low methane concentrations (e.g. 30 nm h@1 for a CH4 concentration of 0.05%) [4]. A linear relation between growth rate and methane concentration can be
observed for concentrations lower than 0.25%. This can be explained by the decrease of activated species (such as atomic hydrogen and methyl radical) and the increase of etching since the relative concentration of atomic hydrogen increases with decreasing methane concentration [6]. In order to increase the growth rate for high quality diamond, deposition at low methane concentration under atmospheric pressure has been attempted. Thereby, a high growth rate of 0.18 mm h@1 has been achieved at a methane concentration of 0.03%, which is an increase by a factor of about six compared to low pressure conditions [5]. A reasonable explanation is the competition between the generation and recombination of atomic hydrogen because the absolute generation of atomic hydrogen decreases with the total pressure in the chamber (see Table 1). SEM micrographs of diamond particles deposited on Si substrates under various CH4 concentrations are shown in Fig. 3. The total growth time and the residence time were fixed at 3 h and 1.7 s for all samples, respectively. At low methane concentration single crystals without any facets can be observed (Fig. 3(a)). In contrast, twin structures can be observed on the diamond particles grown at high methane concentrations as shown in Fig. 3(b)–(d). At low methane concentration, the increase in the ratio between
452
K.K. Hirakuri et al. / Vacuum 63 (2001) 449–454
atomic hydrogen and methyl radical enhanced the crystalline property in the diamond grains. In order to compare the quality of the diamonds grown at different conditions, micro Raman spectroscopy was applied. Fig. 4 shows the Raman Table 1 Growth conditions for CVD diamond deposition under atmospheric pressure Parameters
Range
Substrate material Pressure (kPa) Residence time (s) Methane concentration (%) Deposition time (h) Substrate temperature (K)
Si (1 0 0) 5B200 0.02B5 0.03, 0.1, 0.3, 0.5, 1.0, 1.5, 2.0 3 1120
spectra of diamond grains produced at methane concentrations between 0.1% and 1.0%. In each case the grains were identified as diamond due to the sharp peak at 1332 cm@1. Significant peaks at 1350 cm@1 (D band) and 1530 cm@1 (G band) originating from disordered and glassy amorphous carbon around the grains cannot be observed in the Raman spectra. The full width at half maximum (FWHM) of the peaks was used to characterize the quality of the diamond grains. Typical FWHMs of the diamond peak at 1332 cm@1 are listed in Table 2. A significant increase of the FWHM with increasing methane concentration can be observed. In particular, a strong decrease of the FWHM can be found for a methane concentration of 0.5%.
Fig. 3. SEM images showing the morphology of diamond grains grown on a silicon substrate. Methane concentration: (a) 0.1%, (b) 0.3%, (c) 0.5%, and (d) 1.0%. The total deposition time was 3 h in all the cases.
453
K.K. Hirakuri et al. / Vacuum 63 (2001) 449–454
Fig. 4. Raman spectra of CVD diamond grains grown at different methane concentrations. Table 2 FWHMs of Raman peaks at 1333 cm@1 recorded on diamond grains grown at different methane concentrations Methane concentration (%)
FWHM (cm@1)
1.0 0.5 0.3 0.1
11.53 11.42 10.28 7.79
This result is similar to the dependence observed for the growth rate as a function of methane concentration as shown in Fig. 2. At low methane concentrations, the comparative ratio between atomic hydrogen and methyl radical increases with decreasing methane concentration. Thus, non-diamond species can be etched off, which is also supported by the SEM micrographs in Fig. 3. Next, the quality of diamond grains before and after etching was investigated by Raman spectroscopy since non-diamond components can be identified by this technique. All the spectra have been recorded on the same individual diamond grain by focussing the laser beam at the center of the grain (Fig. 5). In the spectrum of the asdeposited sample (at a methane concentration of 0.5%), besides the dominant diamond peak at 1332 cm@1, a weak peak around 1450 cm@1 can be observed, which indicates the existence of amorphous carbon [7,8]. After the etching treatment the amorphous carbon peak is strongly
Fig. 5. Raman spectra of diamond grains before and after etching. Etching I: residence time=0.5 s, etching II: residence time=1.5 s.
Table 3 FWHMs of Raman peaks at 1333 cm@1 recorded on diamond grains before and after etching
Reference After etching I After etching II
Residence time (s)
FWHM (cm@1)
As deposited 0.5 1.5
11.23 6.58 5.76
reduced. Moreover, the FWHMs for the diamond peaks of the as-deposited and the etched samples were determined as well (Table 3). A reduction of the FWHM by the etching treatment from 11.28 to 6.58 cm@1 can be observed. Furthermore, the residence time of the etching gas also influences the FWHM of the etched samples. While an FWHM of 6.58 cm@1 is observed for the sample etched with a residence time of 0.5 s (denoted as etching I in Fig. 5), the respective value for a residence time of 1.5 s (etching II in Fig. 5) is 5.76 cm@1. This can be explained by the increased generation of atomic hydrogen through extension of the residence time, enhancing the removal of non-diamond components [3]. 4. Conclusions Using the HF–CVD technique, the influence of the methane concentration on diamond growth
454
K.K. Hirakuri et al. / Vacuum 63 (2001) 449–454
was investigated under atmospheric pressure. A decreasing growth rate of the diamond could be observed with decreasing methane concentration. A variation of the CH4 concentration from 0.03% to 1.0% leads to an increase in growth rate by a factor of about 10. On the other hand, the quality of the diamond decreases at the same time. However, the significant enhancement of the growth rate at atmospheric pressure allows production of high quality diamond at low CH4 concentration and still with reasonable deposition rates, making the technique useful for industrial applications.
Acknowledgements The authors would like to thank Mr. T. Toya for performing SEM analysis. This work has been
partially supported by the Center for Research of Tokyo Denki University. References [1] Rye RR. J Appl Phys 1994;76:1220–7. [2] Connell LL, Fleming JW, Chu H-N, Vestyck Jr DJ, Jansen E, Butler JE. J Appl Phys 1995;78:3622–30. [3] Kobayashi T, Hirakuri KK, Mutsukura N, Machi Y. Diamond Related Mater 1999;8:1057–60. [4] Takeuchi D, Yamanaka S, Watanabe H, Sawada S, Ichinose H, Okushi H, Kajimura K. Diamond Related Mater 1999;8:1046–9. [5] Takeuchi D, Watanabe H, Yamanaka S, Okushi H, Kajimura K. Phys Stat Sol 1999;174:101–15. [6] Hayashi K, Yamanaka S, Watanabe H, Sekiguchi T, Okushi H, Kajimura K. J Cryst Growth 1998;183:338–46. [7] Hirakuri KK, Yoshii M, Friedbacher G, Grasserbauer M. Diamond Related Mater 1997;6:1031–5. [8] Leeds SM, Davis TJ, May PW, Pickard CDO, Ashfold MNR. Diamond Related Mater 1998;7:233–7.