Plasma Diagnosis for Microwave ECR Plasma Enhanced Sputtering Deposition of DLC Films

Transcription

1 Plasma Science and Technology, Vol.14, No.2, Feb Plasma Diagnosis for Microwave ECR Plasma Enhanced Sputtering Deposition of DLC Films PANG Jianhua ( ) 1, LU Wenqi ( ) 1, XIN Yu ( ) 2, WANG Hanghang ( ) 1, HE Jia ( ) 1, XU Jun ( ) 1 1 Key Laboratory of Materials Modification by Laser, Ion and Electron Beams (Dalian University of Technology), Ministry of Education, Dalian , China 2 School of Physics and Optoelectronic Technology, Soochow University, Suzhou , China Abstract Application of the Langmuir probe in plasma circumstance for deposition of diamond-like carbon (DLC) thin films usually faces the problem of rapid failure of the probe due to surface insulative coating. In this paper, we circumvent the problem by using the floating harmonic probe technique. In the real circumstance of DLC film deposition, the floating harmonic probe worked reliably over 3 hours, correctly indicating the ion density and electron temperature. The technique was practically used to measure the ion density and electron temperature in DLC film deposition processes using the microwave ECR plasma enhanced sputtering. Combined with the Raman spectroscopic characterization of the films, the conditions for deposition of DLC films were investigated. Keywords: floating harmonic probe, DLC films, microwave ECR plasma, Raman spectroscopy PACS: m, Dq DOI: / /14/2/17 1 Introduction With many excellent properties, diamond-like carbon (DLC) films have drawn much interest and been used in a wide range of contemporary applications [1]. However, in contrast to applied researches, plasma diagnosis for deposition of DLC films seems to be seldom reported, though it is important for monitoring the process. A possible reason is that the Langmuir probe, which has been conveniently used in non-polluting discharges, can not be applied in the circumstances of DLC film depositions because of rapid failure of the probe due to insulative surface coating on the probe tip [2 4]. Efforts have been made to overcome the weakness of Langmuir probes. OLIVER and CLEMENTS had developed the radio frequency floating double probe [5]. They operated the double probe in the megahertz range and gave the electron density and the electron temperature by mersuring the sheath capacitance and the resistance of the probe. GODYAK and co-workers developed a RF floating-type probe using the nonlinear characteristics of the probe sheath in RF plasma [6]. The floating-type probe has been used to measure the electron temperature for a fast time resolution in the Tokamak plasma [7 9]. In the work of Min-Hyong LEE et al., they developed a floating-type probe and used coated probe tip in CF 4 plasma [10]. In our work, we applied the floating harmonic probe to diagnose the plasma in deposition of DLC films using the microwave ECR plasma enhanced sputtering. The floating probe gave consistent results with those obtained from single Langmuir probe in non-polluting Ar discharge, and worked steadily in the real case of DLC film deposition where the Langmuir probe cannot work. The relation between the plasma parameters and the film properties was investigated. 2 Principles of the floating harmonic probe The construction of the floating harmonic probe is schematically shown in Fig. 1. It consists of a metal probe, a capacitance, a resistor and a signal generator. When working, alternative bias voltage of several kilohertz was applied on the probe, and the probe current Fig.1 A schematic diagram of the floating harmonic probe

2 PANG Jianhua et al.: Plasma Diagnosis for Microwave ECR Plasma Enhanced Sputtering Deposition was detected by an oscilloscope through a sampling resistor. In addition, the amplitudes of the harmonics of the current signal are obtained by fast Fourier transform (FFT) for later analysis of the plasma parameters. Assuming that the electrons are of a Maxwellian distribution, the current through the probe tip, whose bias voltage is V B, can be written with sum of the electron and the ion current [11] i pr = i is i es exp[e(v B V p )/T e ], (1) where, T e is the electron temperature, V p is the plasma potential, i is is the ion saturation currents, and i es are the electron saturation currents. Van NIEUWENHOVE et al. [8] had proved that i pr can be expanded with modified Bessel function I k (z) as: i pr = i is i es exp [ e ( ( ) )/ ] ev0 V V p Te I0 T e 2i es exp [ e ( )/ ] ( ) ev0 V V p Te I k cos (kωt). T e k=1 (2) If a capacitor is used to block the dc current, the DC component of the i pr is zero, and the probe current becomes i pr = 2i es exp[e(v V p )/T e ][I 1 cos(ωt) +I 2 cos(2ωt) +...] (3) = 2i es exp[e(v V p )/T e ](i 1ω + i 2ω +...). From Eq. (3), we know that the amplitudes of each harmonic current are functions of T e and V 0 only, thus the electron temperature may be obtained by comparing amplitudes of any two harmonics. The ratio of the first and second harmonics amplitudes becomes as follows i 1ω / i 2ω = I 1 (ev 0 /T e )/I 2 (ev 0 /T e ). (4) Solving the Bessel equation, the electron temperature may be calculated with measured amplitudes of 1ω and 2ω. On the other hand, the first harmonic of the probe current is given as i 1ω = 2i is (I 1 /I 0 )cos(ωt) = 2(0.61en i u B A)(I 1 /I 0 )cos(ωt). (5) Thus from Eq. (5) the ion density n i may also be obtained with the measured amplitude of the first harmonic of the probe current. 3 Experimental details Depositions were performed in an ECR sputtering system onto Si (100) substrates, the Si substrates had been cleaned ultrasonically in acetone and ethanol before being put into the chamber. Fig. 2 shows a schematic diagram of the microwave ECR plasma enhanced sputtering setup. Microwave (2.45 GHz) power was fed into the chamber through a rectangular wave guide and a quartz pressure window. Electric coils were arranged around the chamber to achieve proper magnetic field (875 G) for ECR discharge. A ring-shaped sputtering carbon target designed for surface coating of cylindrical workpieces was placed adjacent to the ECR source, co-axial with it. Pure Ar gas was used for sustaining the discharge and sputtering the target. A negative voltage was applied to the carbon target with DC power supply. With the high density ECR plasma, a satisfactory sputtering rate may be achieved. The conditions for deposition are summarized in Table 1. Fig.2 A schematic diagram of MW-ECR plasma enhanced unbalanced magnetron sputtering deposition system Item V p p Table 1. Probe area Resistor Sputter gas Target Substrate Temperature Gas pressure Microwave power Target voltage Time Experimental conditions Condition 5 V m kω Ar Carbon Si(100) Room temperature Pa W 400 V 1 hour All diagnostic experiments were executed with the probes being set close to the axial center where the substrate locates when depositing. The probe tip in Fig. 1 is 7 mm in length and 0.4 mm in diameter, made of copper, and a 5V p p sinusoidal signal was applied between the probe tip and the ground. The current was calculated from the measured voltage drop across the current sensing resistor. A Labview program was worked out to analyze the harmonics of the current signal and make 173

3 Plasma Science and Technology, Vol.14, No.2, Feb calculation automatically to give the electron temperature and ion density. In addition, a single Langmuir probe (SLP) was placed in the vicinity of the floating probe, measuring simultaneously for comparison. 4 Results and discussion Fig. 3 shows the electron temperature and ion densities obtained by the floating harmonic probe and single Langmuir probe in 0.5 Pa pure Ar discharge at various microwave powers between 200 W to 800 W. It can be seen that good agreement was obtained between the results obtained by floating harmonic probe and single langmuir probe, indicating reliability of the floating harmonic probe measurements. Fig.3 Comparison of the electron temperature and the ion density measured by two types of probe versus the microwave power at 5 mtorr argon pressure (color online) Fig. 4 displays the electron temperature and ion density, measured by the floating harmonic probe, with the typical deposition conditions of 400 V DC voltage, 0.5 Pa pressure and 400 W microwave power. It suggests stable work of the probe during the entire deposition period up to 3 hours. The probe was checked after the experiment and it was found to be coated with a layer of film and become insulative. It can be seen in Fig. 4 that the electron temperature changed little but the ion density increased continuously with time. The trend for the ion density is contrary to what expected. Generally, the measured density will decrease, though insignificant, after the probe being used in the deposition circumstance for several hours. This is caused by 174 an insulative coating of the probe which reduces the effective capacitance between the probe and the plasma, and thus reduces the probe current and correspondingly the ion density calculated [10]. To investigate the reason for the abnormal increasing of the ion density with time, we measured the electron temperature and ion density using a Langmuir probe in the same condition except for sputtering the target. The results are shown in Fig. 5. Both the electron temperature and ion density in Fig. 5 changed similarly as compared to those shown in Fig. 4, suggesting the real increasing of the ion density with time, and the results of the floating harmonic probe may be justified. The increasing of ion density is possibly due to the unstable output of the microwave power source using a magnetron [12]. Fig.4 The electron temperature and ion density measured by floating-probe versus deposited time The floating harmonic probe diagnosis was then combined with the deposition of DLC films to investigate the relation between the plasma condition and the film properties. The films were deposited at different pressures of 0.17 Pa, 0.5 Pa and 1 Pa, and the electron temperature and ion density were measured simultaneously. No bias was applied to the substrate when coating. The measured electron temperature and ion density are shown in Fig. 6, and the deposited films were characterized by Raman spectroscopy as shown in Fig. 7. Information extracted from the spectra in Fig. 7 is listed in Table 2. The spectra shown in Fig. 7 are not typical for the DLC film which consists of a broad G bands and a low D bands [13]. Nevertheless, some sp3 bondings should have formed in the films as judged by the lower maximum of the D bands. We considered

4 PANG Jianhua et al.: Plasma Diagnosis for Microwave ECR Plasma Enhanced Sputtering Deposition the ion flux to the substrate is smaller due to the lower ion density, at the higher pressure of 1 Pa the ion energy is reduced due to enhanced collision and scattering of ion in the sheath, both result in reduction of total bombardment energy to the substrate. Fig.5 The electron temperature and ion density measured by SLP in discharge generated merely by microwave Fig.7 Raman spectra of hydrogen-free DLC films deposited at (a) 0.17 Pa, (b) 0.5 Pa and (c) 1 Pa, respectively (color online) 5 Fig.6 The electron temperature and ion density measured by floating-type probe versus the Argon pressure that the non-typical DLC feature of the films is due to insufficient ion bombardment on the films (note that no bias was applied to the substrate) and insufficient ionization of carbon atoms. An evidence is that the film deposited at 0.5 Pa showed the lowest I(D)/I(G) ratio, which means the highest sp3 content [14,15]. It may be interpreted as: while at the lower pressure of 0.17 Pa Conclusions A floating harmonic probe was established for measuring the electron temperature and ion density in plasma circumstances for insulating films deposition. Good agreement was obtained between the results from floating harmonic probe and single Langmuir probe in pure Ar discharge, indicating reliability of the floating harmonic probe measurements. The floating harmonic probe worked stably, gave the ion density and electron temperature in the whole period of DLC film deposition with microwave ECR plasma enhanced sputtering. Combined with the Raman spectroscopic characterization of the films, the conditions for deposition of DLC films were investigated. 175

5 Plasma Science and Technology, Vol.14, No.2, Feb Table 2. Analysis of Raman spectra through Gaussian fitting Pressure D peak G peak I D/I G Area Center FWHM Height Area Center FWHM Height (Pa) (cm 1 ) (cm 1 ) (cm 1 ) (cm 1 ) References 1 Ru Lili, Huang Jianjun, Gao Liang, et al. 2010, Plasma Science and Technology, 12: Chen F F. 1968, Plasma Diagnostics Technique, edited by Huddlestone R H. Academic Press, New York 3 Langmuir, Mott-Smith H. 1924, General Electric Reviews, 27: Boedo J A, Gray D S, Chousal L, et al. 1998, Review of Scientific Instruments, 69: Oliver B M, Clements R M. 1970, Journal of Applied Physics, 41: Godyak V A, Kuzovnikov A A, Khadir M. 1968, Radiotekhnika i elektronika, 3: Rudakov D L, Boedo J A, Moyer R A, et al. 2001, Review of Scientific Instruments, 72: Nieuwehove R V, Oost G V. 1988, Review of Scientific Instruments, 59: Boedo J A, Gray D, Conn R W, et al. 1999, Review of Scientific Instruments, 70: Lee M H, Jang S H, Chung C W. 2007, Journal of Applied Physics, 101: Liebermann M A, Lichtenberg A J. 1994, Principles of Plasma Discharges and Material Processing. Wiley, New York 12 Shu Xiangshen, Wu Qinchong. 2000, Vacuum Electronics, 1: 41 (in Chinese) 13 Dilou R O, Woolam J A, Katkanant V. 1984, Physical Review B, 29: Robertson J, O Reilly E P. 1987, Physical Review B, 35: 2946; Robertson J. 1986, Advances in Physics, 35: Ferrari A C, Robertson J. 2000, Physical Review B, 61: (Manuscript received 6 August 2011) (Manuscript accepted 20 December 2011) address of corresponding author XU Jun: xujun@dlut.edu.cn 176