Subscribe Now Subscribe Today
Research Article
 

Silicon Carbon Nitride Thin Films Deposited by Pulsed Microwave Plasma Assisted Chemical Vapour Deposition



Paul Kouakou, Pamela Yoboue, Bafetigue Ouattara, Mohammed Belmahi and Jamal Bougdira
 
Facebook Twitter Digg Reddit Linkedin StumbleUpon E-mail
ABSTRACT

Objective: This study aimed to synthesize crystalline silicon carbon nitride thin films by pulsed microwave plasma assisted chemical vapour deposition in N2/CH4 gas mixture on silicon substrates. Methodology: Prior to deposition, the pulsed microwave discharges were analyzed in situ by Time Resolved Optical Emission Spectroscopy (TROES). The behavior of the emissive radicals present in the plasma is observed during the discharge and post-discharge phases. The study of the pulsed plasma shows that, in the post-discharge, CN and C2 radicals continue their emission and fluorescence was observed which is explained by the energy transfer from the metastable states of N2 to these species. The films were realized according to the pulse parameters and were then characterized by Scanning Electron Microscopy (SEM) and Transmission Electron Microscopy (TEM), in order to observe their morphology and structures. Results: These images showed that films have nano-crystalline structure. The films chemical composition and their bonding structure are also analyzed by X-ray Photoelectrons Spectroscopy (XPS). The films contain C, N, Si and a few quantity of O2 coming from their exposure to atmosphere after deposition. The silicon in the films comes from the etching of the silicon substrate. The observation of C1s, N1s and Si2p levels showed the presence of C-N, Si-C and Si-N bonds. Conclusion: The pulsed microwave mode is very advantageous for the molecules dissociation compared to the continuous mode. It permitted us to follow the growth and etching kinetic in order to control the films chemical composition. Thus, we obtained 16% of nitrogen in the film with a post-discharge duration of 3 msec.

Services
Related Articles in ASCI
Similar Articles in this Journal
Search in Google Scholar
View Citation
Report Citation

 
  How to cite this article:

Paul Kouakou, Pamela Yoboue, Bafetigue Ouattara, Mohammed Belmahi and Jamal Bougdira, 2017. Silicon Carbon Nitride Thin Films Deposited by Pulsed Microwave Plasma Assisted Chemical Vapour Deposition. Journal of Applied Sciences, 17: 306-314.

DOI: 10.3923/jas.2017.306.314

URL: https://scialert.net/abstract/?doi=jas.2017.306.314
 
Received: December 31, 2016; Accepted: March 11, 2017; Published: May 15, 2017


Copyright: © 2017. This is an open access article distributed under the terms of the creative commons attribution License, which permits unrestricted use, distribution and reproduction in any medium, provided the original author and source are credited.

INTRODUCTION

Liu and Cohen1 predicted a hardness of β-C3N4 of 427 GPa close to that of diamond (443 GPa), carbon nitrides films are considered as a very promising materials for mechanical and medical applications, due to their hardness and low friction coefficient, especially as a protective layer against corrosion, oxidation, stripe, wear and fatigue2.

Since the beginning of this study, many studies have been done with many techniques but the synthesis of a mono-phase β-C3N4 is not yet achieved. For most of studies, it is impossible to control the films phases, their stoichiometry and especially their nitrogen content.

Microwave Plasma Assisted Chemical Vapour Deposition (MPACVD) was one of the most used techniques to obtain crystalline CNx phases, to the high gas temperature and dissociation rates, as it has been proved for diamond synthesis3-6. In addition, the possibility to increase the quality and the growth rate of diamond films by pulsed microwave plasma has been proved by other researchers7,8. Indeed, pulsed microwave assisted chemical vapour deposition is a way that permits to increase the injected microwave power without increasing the mean substrate temperature. It thus allows increasing the dissociation rate of the molecules and the density of the atomic and radical species without really increasing the average gas temperature and so the substrate temperature.

In this study, the result of the in situ diagnostic of the pulse microwave plasma by TROES according to the mean pulse parameters (frequency, duty cycle and post-discharge duration) is presented. Films were also realized and characterized according to these pulse parameters in order to control the reactivity kinetic. By varying the pulse and discharge parameters, the nitrogen incorporation in the film can be controlled and a correlation between the nitrogen content in the film and the species present in the plasma can be made.

Scanning electron microscopy (MEB) is used to characterize the films morphology and X-ray Photoelectrons Spectroscopy (XPS) for the bonding structure and chemical composition.

MATERIALS AND METHODS

The silicon carbon nitride thin films were synthesized in a reactor presented in Fig. 1. The main part of the experimental setup has been described by Lamara et al.7. For the pulsed microwave plasma study, a low frequency function generator which permits to pulse the microwave generator is added to the reactor used for the continuous mode. The main parameters of the microwave generator pulse (the frequency (F), the duty cycle (a), the discharge duration (Td) and the post-discharge duration (Tpd)) were varied as mentioned in Table 1. The TROES system used for pulsed discharge analyses was explained by Kouakou et al.9 and Kouakou10.

Fig. 1:Experimental set up for film deposition by pulsed microwave

Table 1:Complete microwave pulses parameters for the TROES and films deposition as a function of the duty cycle, the frequency and the post-discharge duration
Td: Discharge duration, Tpd: Post-discharge duration

The gas used for the depositions was composed of a mixture of N2 and CH4. In this study, the CH4 percentage was fixed at 4 %. The substrate was Si (100). The gas pressure was 5 kPa and the total gas flow rate was 50 sccm. For the film deposition, the substrate temperature was around 900°C. The deposition duration of all the films was 8 h. These parameters were chosen according to a previous study which has been done by MPACVD in a continuous mode9.

Scanning Electron Microscopy (SEM) observations were performed on a JEOL JSM-6500F using an accelerating voltage of 5 kV to analyze the general morphology of the films.

The X-ray Photoelectrons Spectroscopy (XPS) analyses were also done in order to determine the bonding structure and the surface chemical composition. The XPS spectra were obtained by a multi-detection analyzer (VSW CL150 in fixed analyzer transmission mode). The unmonochromatized Mg Kα (1253.6 eV) source operated at 12 kV and 10 mA. The pressure in the analytical chamber was 10‾7 Pa. The resolution was lower than 1 eV. The charge effects were corrected using Ag3d5/2 at 368.2 eV11. The recorded lines (Si2p, N1s, C1s and O1s) were fitted using a curve-fitting program with Gaussian-Lorentzian (XPSPEAK4.1).

RESULTS AND DISCUSSION

Diagnosis of the pulsed plasma: Time Resolved Optical Emission Spectroscopy (TROES) was a non-intrusive in situ technique that allowed the study of the creation and the loose of the emissive species in the discharge and post-discharge duration. It permitted to understand the temporal physicochemical reactions and to control the creation and loss processes of the species in the plasma that can help to know relatively which specie contributes to the deposition and the etching during the discharge and the post-discharge phases. The film deposition being a competition between growth and etching, it is necessary to vary the pulse parameters in order to increase the quantity of the species that contribute to the deposition and to decrease those that are responsible of the etching. Figure 2 shows an overlay of a typical function generator signal and the OES intensity emitted by CN as observed on an oscilloscope.

Figure 2 shows that the emission intensity of the radical didn’t follow all the time the shape of the function generator. When the function generator was switched on, there was an abrupt increasing of the radical emission intensity in a few microseconds and then it continuous to increase but progressively. The first step was attributed to electronic excitation effect. The second one was probably an equilibrium establishment between excitation and production of the radical. The inverse phenomenon appears when the microwave generator was switched off. During the post-discharge phase, after the abrupt decreasing of the emission intensity due to the end of the electronic effect, the fluorescence of the species continues until a few microseconds. This emission of the radicals without an external energy source was probably due to metastable species. Indeed, the metastable states have a long life time duration (from a few milliseconds to a second) compared to the normal excited state (a few nanoseconds).

Fig. 2:Function generator signal and the OES intensity emitted by CN as observed on the oscilloscope
 
The mean microwave power is fixed at 1000 W, the frequency is 500 Hz and the duty cycle is 50%. The CH4 percentage is 4%, the pressure is 50 mbar and the total flow rate is 50 sccm

The metastable can continue to transfer excitation to the radicals created during the discharge phase in order to excite them in the post-discharge phase. These behaviors of the radicals in pulsed plasma have been observed by other researchers7,12,13. In the case of a mixture of N2/CO2/CH4 gas mixture, Bougdira et al.12 observed a gradual increasing in the intensity emitted by CN while CH and Ha reach their steady states after 40 msec of discharge duration. Researchers thought that the excitation of CH and Ha occur via a direct dissociation excitation process of CH4 while CN requires a stage of formation by recombination between atomic nitrogen and carbon radicals before being excited. In the case of a gas mixture H2/CH4, Lamara et al.7 observed an immediate stabilization of the emission intensity of hydrogen, but not that of C2 and CH. Researchers also explained this by the fact that hydrogen is excited electronically while the formation and excitation of CH and C2 require longer time. In this study, CN and C2 are probably formed by the same way as explained by Bougdira et al.12 and Lamara et al.7. There are results of many dissociation and recombination processes. Another reason for the emission of the species in the post-discharge duration can be due to the excitation by molecular vibrations. Indeed, Guerra et al.14 shown that in nitrogen plasma, the energy transfer by nitrogen vibrational states was important than the electronic effect when the pressure in the reactor chamber exceed 1 τ (1.33 mbar). The evolution of the emission intensity of the CN and C2 radicals in the discharge and post-discharge phases has been studied by TROES as a function of the pulse parameters and presented in the Fig. 3. The emission intensities have been normalized out of the maximum for each line.

At a constant average microwave power, by decreasing the duty cycle of the pulse, we increase the peak microwave power (Table 1). That involves an increasing of the dissociation rate of the molecules and consequently an increasing of the formation of the radicals. Indeed, at low duty cycle, the dissociation probability of the molecules increases due to the increasing of the microwave energy. Hence, the plasma kinetic was changed. The atomic nitrogen and probably atomic carbon densities increase and thus the formation of the CN and C2 radicals increase during the discharge when the duty cycle decreases. The slope of radical emission intensity increases in the discharge, when the duty cycle decreases. For a high duty cycle, the microwave energy was low and then an equilibrium state between the excitation, the creation and the loss of the species is quickly reached. The density of the created species will probably not increase because the peak power was not sufficient in that case. To study the pulse frequency effect, the pulse duty cycle was kept constant at 50%. Increasing the pulse frequency means to decrease the period duration and so the duration of the discharge and that of the post-discharge. So, it was necessary to get the best compromise in order to get more time to deposit and less time for etching. Figure 3b shows that, the time has been normalized out of the period for each line in order to have the same shape for all pulse frequency.

Fig. 3(a-c):
Evolution of the TROES emission intensity of the CN and C2 radicals in the discharge and post-discharge as a function of (a-b) Duty cycle, (c-d) Frequency and (e-f) Post-discharge duration

Figure 3 shows that when the pulse frequency decreases, the share of radicals continues to emit in the post-discharge decreases. The amount of metastables, which are able to transfer energy to the other radicals, decreases when the pulse frequency decreases that means, when the duration of the post-discharge increases. During the study of the effect of the post discharge duration, the discharge duration has been fixed at 1 msec. In the post-discharge phase, the microwave was turned off, so there was no external source of energy for molecules dissociation or species creation. Table 1 shows that, when the post-discharge duration increases, for a constant average microwave power, the duty cycle and the frequency decrease. Thus the crest power increases. Varying the duration of the post-discharge also means to modify the duration of actions of the plasma species on the film etching and deposition processes.

Fig. 4(a-e): SEM micrographs of the films deposited in pulsed mode at different frequencies, (a) 100 Hz, (b) 500 Hz, (c) 600 Hz, (d) 700 Hz and (e) 800 Hz

Fig. 5(a-c): SEM of silicon carbon nitride films deposited at different post-discharge duration Tpd, (a) 0.2 msec, (b) 2 msec and (c) 3 msec

In both cases, the magnitude of the jump of the CN and C2 emission intensity at the beginning of the discharge was equal to the drop at the beginning of the post-discharge. That means that the processes at the beginning of the discharge and the post-discharge have probably the same source. They were due to the electronic effect which appears immediately when the microwave power was switched on and disappeared as soon as it is stopped. Figure 3 also shows that the share of the electronic excitation of the radical C2 is much less significant compared to that of CN. The steady state was reached sooner for CN than for C2. It means that the formation of the C2 radical take longer than CN.

Morphological characterization of the obtained films: The SiCN films were realized and characterized according to the plasma pulse parameters. The general morphology of the surface of the films synthesized according to the pulse frequency was observed by SEM as shown in Fig. 4.

The films obtained by varying the post-discharge duration have also been observed by SEM as shown in Fig. 5. The micrographs show that the quality of the film deteriorates when the post-discharge duration increases. As mentioned in the plasma study, increasing the post-discharge duration means to let longer time to the species to the etching.

Fig. 6(a-i):
Chemical composition obtained by XPS as a function of the microwave pulse parameters (Table 1) one a film deposited at 4% CH4, 50 mbar, 50 sccm. Evolution of the C, N and Si percentages as a function of (a-c) Duty cycle, (d-f) Frequency and (g-i) Post-discharge duration. A comparison is done with a film deposited in the continuous mode at 4% CH4, 1000 W, 50 mbar, 50 sccm

Chemical composition and bonding structure: The film composition was analyzed by XPS without removing the natural oxidation after exposition to air in order to avoid formation of carbon bonds. Indeed, the bombardment of the film by argon ions modifies the bonding structure of the film. So, in this study all the films were observed without any surface cleaning. Consequently the oxygen has been substrate to the calculation of the general composition. The silicon composition coming from the Si-Si bonds have been substrate because this part probably comes from the substrate without any recombination with the other elements. Indeed, the film was thin and porous, thus the x-ray can reach the substrate silicon.

Figure 6 presents the chemical composition obtained from XPS of the films deposited as a function of the microwave pulse parameters.

In Fig. 6, the XPS shows that the nitrogen percentage in the film first increases when the duty cycle increases until 40-50%, then it decreases abruptly. The nitrogen percentage increases as the pulse frequency increases and reaches a maximum value around 10% for a pulse frequency of 600 Hz. By continuing to increase the pulse frequency, the nitrogen percentage in the films decreases. The silicon percentage behavior was exactly the contrary of the nitrogen one. The carbon percentage decreases continuously when the frequency decreases. When the post-discharge duration increases, the nitrogen percentage in the film increases. For the post-discharge duration of 3 msec, the nitrogen percentage reaches 16%. The carbon percentage in the film decreases when the post-discharge duration decreases until 2 msec and then increases. The silicon percentage behavior was opposite of that of carbon.

During the post-discharge duration, the atomic nitrogen created during the discharge phase contributes principally to the etching of the film. When the post-discharge duration increases, probably the density of the species that contribute to the growth of the film decreases and those that enhance the etching increases. The dissociation rate was also increased at high microwave crest power when the post-discharge duration increases and favors the creation of the atomic N and H species that etched the film. The species, which participate to the growth were formed during the plasma on-duration. The atomic nitrogen created during the plasma-on time which was not used to form deposition radicals was used to etch the film during the plasma-off time and a part incorporates the film. So, during the plasma-off time, there was a competition between nitrogen incorporation and film etching by atomic nitrogen. In the case of diamond growing by pulsed microwave H2/CH4 plasma, Lamara et al.15 have shown that the film growing operates principally during the plasma on-time. In that case, atomic hydrogen has been identified as the non-diamond carbon etching agent. So, increasing atomic hydrogen creation means increasing etching of the non-diamond phase etching. The difficulty with carbon nitride study is the lack of precise knowledge about the species that contribute to the growth and those which are responsible of the film etching. This study showed that the nitrogen species produced in the discharge are probably the responsible of the film etching (etching of C and Si) during the post-discharge phase. In fact, Bulou et al.16 observed the presence of atomic nitrogen, atomic silicon and atomic carbon in such a discharge by OES. Nitrogen percentage in the film was also increased with the post-discharge duration. The film etching observed by SEM was confirmed by the XPS results. In fact, the Fig. 6 shows that for all pulse frequency, the silicon percentage observed in the film was greater than that observe in the film deposited with the continuous mode. The substrate was much more etched when the discharge was pulsed. The film surface composition cannot evolve correctly due to the growth mechanism as explained in detail elsewhere17. This growth mechanism was responsible of a changing of the kinetic at the surface after obtaining a continuous film. So, the XPS results must be taken with care because they were somewhat disturbed by this growth mechanism. The results were probably not homogeneous throughout the growth so the composition at the surface was probably lightly different from that near the interface film-substrate.

CONCLUSION

Nano-crystalline silicon carbon nitride films were synthesized by PMPACVD in CH4 and N2 gas mixture on silicon substrate. The SEM and TEM showed that the films were crystalline. The morphology of the films changes slightly while changing the pulse parameters. High pulse frequency and high post-discharge duration involved deterioration of the film due to the etching phenomenon. The XPS analyses proved that the chemical composition of the film depend on the parameters of the pulse of the plasma. The nitrogen contained in the film increases when the pot-discharge duration increases. The XPS study showed the presence of some covalent bonding (CN, SiN) corresponding to the C3N4 and the SiCN phases.

SIGNIFICANT STATEMENTS

The pulsed microwave mode is very advantageous for the molecules dissociation
The films chemical composition is well controlled by controlling the plasma composition
Microwave plasma assisted CVD technique in CH4 and N2 gas mixture on silicon substrate permitted to have crystallized SiCN films

ACKNOWLEDGMENTS

This study was supported by the "Fonds National de la Recherche" of Luxembourg. This study was also partly supported by the "Ministère délégué à l’Enseignement Supérieur et à la Recherche" of France and the Ivory Coast government. The authors wish to thank J. Ghanbaja, B. Assouar, from Université de Lorraine for TEM and SEM analyses respectively.

REFERENCES
Bertran, E., F.J. Pino, G. Viera and J.L. Andujar, 2002. Hard coatings for mechanical applications. Vacuum, 64: 181-190.
CrossRef  |  Direct Link  |  

Bougdira, J., M. Remy, P. Alnot, C. Bruch, J.K. Kruger, H. Chatei and J. Derkaoui, 1998. Combined effect of nitrogen and pulsed microwave plasma on diamond growth using CH4-CO2 gas mixture. Thin Solid Films, 325: 7-13.
CrossRef  |  Direct Link  |  

Bulou, S., L. Le Brizoual, R. Hugon, L. De Poucques, M. Belmahi, H.N. Migeon and J. Bougdira, 2009. Characterization of a N2/CH4 microwave plasma with a solid additive Si source used for SiCN Deposition. Plasma Process. Polym., 6: S576-S581.
CrossRef  |  Direct Link  |  

Chatei, H., J. Bougdira, M. Remy and P. Alnot, 1999. Optical emission diagnostics of permanent and pulsed microwave discharges in H2-CH4-N2 for diamond deposition. Surf. Coat. Technol., 116: 1233-1237.
CrossRef  |  Direct Link  |  

Chatei, H., J. Bougdira, M. Remy, P. Alnot, C. Bruch and J.K. Kruger, 1997. Effect of nitrogen concentration on plasma reactivity and diamond growth in a H2-CH4-N2 microwave discharge. Diamond Related Mater., 6: 107-117.
CrossRef  |  Direct Link  |  

Chatei, H., J. Bougdira, M. Remy, P. Alnot, C. Bruch and J.K. Kruger, 1997. Combined effect of nitrogen and pulsed microwave plasma on diamond growth. Diamond Related Mater., 6: 50-510.
CrossRef  |  Direct Link  |  

Gajdos, M., A. Eichler and J. Hafner, 2003. Ab initio density functional study of O on the Ag (001) surface. Surf. Sci., 531: 272-286.
CrossRef  |  Direct Link  |  

Gruger, H., D. Selbmann, E. Wolf and A. Leonhardt, 1997. Influence of RF plasma properties on the deposited PACVD CNx layers. Surf. Coat. Technol., 97: 109-113.
CrossRef  |  Direct Link  |  

Guerra, V., P.A. Sa and J. Loureiro, 2004. Kinetic modeling of low-pressure nitrogen discharges and post-discharges. Eur. Phy. J. Applied Phys., 28: 125-152.
CrossRef  |  Direct Link  |  

Kouakou, P., 2008. Plasma physics and optoelectronic. Ph.D. Thesis, Nancy University, France.

Kouakou, P., M. Belmahi, V. Brien, V. Hody, H.N. Migeon and J. Bougdira, 2008. Role of silicon on the growth mechanisms of CNx and SiCN thin films by N2/CH4 microwave plasma assisted chemical vapour deposition. Surf. Coat. Technol., 203: 277-283.
CrossRef  |  Direct Link  |  

Kouakou, P., V. Brien, B. Assouar, V. Hody, M. Belmahi, H.N. Migeon and J. Bougdira, 2007. Preliminary synthesis of carbon nitride thin films by N2/CH4 microwave plasma assisted chemical vapour deposition: Characterization of the discharge and the obtained films. Plasma Process. Polym., 4: S210-S214.
CrossRef  |  Direct Link  |  

Lamara, T., M. Belmahi, G. Henrion, R. Hugon, J. Bougdira, M. Fabry and M. Remy, 2005. Correlation between the characteristics of a pulsed microwave plasma used for diamond growth and the properties of the produced films. Surf. Coat. Technol., 200: 1110-1116.
CrossRef  |  Direct Link  |  

Lamara, T., M. Belmahi, J. Bougdira, F. Benedic, G. Henrion and M. Remy, 2003. Diamond thin film growth by pulsed microwave plasma at high power density in a CH4-H2 gas mixture. Surf. Coat. Technol., 174: 784-789.
CrossRef  |  Direct Link  |  

Liu, A.Y. and M.L. Cohen, 1989. Prediction of new low compressibility solids. Science, 245: 841-843.
CrossRef  |  Direct Link  |  

Sakamoto, Y. and M. Takaya, 2003. Fabrication of nitrogen included carbon materials using microwave plasma CVD. Surf. Coat. Technol., 169: 321-323.
CrossRef  |  Direct Link  |  

Zhang, Y.P. and Y.S. Gu, 2001. Effect of silicon impurity on carbon nitride films prepared by microwave plasma chemical vapor deposition. Mater. Sci. Eng., B, 85: 38-42.
CrossRef  |  Direct Link  |  

©  2020 Science Alert. All Rights Reserved