acta physica slovaca vol. 54 No. 2, 163 - 167
April 2004
STUDY OF PLASMA PROCESSES IN AFTERGLOW BY MEANS OF ELECTRON
SPIN RESONANCE 1
J. Janča2, A. Tálský, V. Kudrle, M. Janča, V. Zvoníček, Z. Frgala
Department of Physical Electronics, Faculty of Science, Masaryk University Kotláfská 2, CZ-61137Bmo, Czech Republic
Received 7 April 2003 in final form 3 October 2003, accepted 14 November 2003
This contribution is intended to summarize in a practically useful form the basic information about the detection of the gas phase atoms and free radicals by electron spin resonance spectroscopy (ESR) in gas discharge afterglow and in remote plasma reactors. This paper shows how ESR could be effectively used for the monitoring of relative or absolute concentrations of gas phase H, O, N atoms during the volume plasma chemical and wall recombination processes.
PACS: 82.33XJ, 52.80Pi, 87.64Hd
1 Introduction
Electron spin/paramagnetic resonance effect is based on resonant absorption of microwave photons by transitions between Zeeman split energy levels. Before entering into discussion of the spectra of the individual species taking part in plasma chemical processes, there are certain remarks, which apply generally that ought to be noted. Nearly all of the transitions of interest involve the absorption of magnetic dipole radiation between Zeeman levels. All of the transitions of interest for present purposes will be governed by the selection rule on the magnetic quantum number AMj — ± 1, so that for absorption of either electric or magnetic dipole radiation a component of the corresponding field vector must be perpendicular to the external magnetic field. With the addition of a nuclear spin I in the strong field limit, J and I are uncoupled from each other and interact separately with the field. Since there are 2 J+1 values of Mj and 21+1 values for M/(AM — 0) the total number of spectral lines for given J and I is 2J(2I + 1). Summary of some atoms and species whose gas phase ESR spectra have been detected experimentally is presented e.g. in [1].
1 Presented at XIVth Symposium on Application of Plasma Processes, Liptovský Mikuláš (Slovakia), January 2003. 2E-mail address: jan92@physics.muni.cz
0323-0465/04 © Institute of Physics, SAS, Bratislava, Slovakia
163
164
J. Janča et al.
2  Determination of rate coefficients of gas phase chemical reaction by ESR
The processes for the deposition of SiC, SKD2 and Si3N4-like films by plasma - enhanced chemical vapor deposition (PECVD) in radiofrequency or microwave plasma reactors are generally used for synthesis of amorphous materials. Kinetics of the reactions H+SiH4, O+TEOS (tetraethoxysilane) and N+HMDSZ (hexamethyldisilazane) were studied in works [2-4].
In the present study the ESR flow tube technique has been applied to characterize the reactant combinations that are of interest in the synthesis of amorphous SKD2 and Si3N4-like materials. Study of reaction kinetics by ESR was described in detail in [1, 5]. Gas phase chemical reactions for deposition of a-Si:H, a-Si:CH, Si02-like and Si3N4-like materials in a remote hydrogen and oxygen plasma reactor have been quantitatively characterized with ESR in [2-7]. A schematic drawing of the experimental setup is presented in Fig. 1. The plasma discharge was sustained in the quartz tube within a cavity (13.56 MHz, 25 W). Pure molecularN2 (O2) was injected into the plasma through a quartz tube that passed, after a right-angle bend, into the ESR cavity (X-band). The density of [N] or [O] atoms in the distance 35 cm from the plasma exciting RF cavity reaches more than 1014 cm~3. Between the two cavities the HMDSZ or TEOS vapors were injected into the flux of partially atomized N2 or O2 via small diameter tube. The total pressure in the ESR cavity cavity during all measurements was maintained at 250 Pa. The ESR spectrometer was calibrated with molecular oxygen as is described in [1]. ESR measurements provided absolute values of [N] or [O] and the correction for the loss of [N] ([O]) by processes that are first order in N (O) and independent of HMDSZ (TEOS), such as wall recombination.
Fig. 1. Apparatus used for determination of O+TEOS and N+HMDSZ kinetics. 1 - ESR cavity, 2 - pumping system, 3 - RF discharge cavity.
Monitoring of N-atom (O-atom) loss as a function of time (residential time) was the basis of this experiment. Correction for the [N] or [O] loss due to the wall recombination is achieved by recording [N] or [O] at each position x downstream from the injector both with ("on") and without ("off") flowing HMDSZ (TEOS). For the case without flowing HMDSZ we can write
Study of plasma processes in afterglow.
165
Fig. 2. Nitrogen atom density in afterglow as a function of O2 admixture concentration. Experiment was carried out at 450 Pa, nitrogen flow was 50 seem. (1 sccm=1.7 x 10~3 Pam3s_1).
for the the N-atom density
d[N]oS/dt = -k2[N]oS => [N]oS[N2]0/[N2]x = Ci exp(-A;2t), (1)
where k2 is the wall recombination coefficient, C\ is a constant depending on experimental arrangement, x is the distance of ESR cavity from plasma exciting RF discharge. From (1) we get for the case with flowing HMDSZ
d[iV]0„/dt   =   -(k2 + kx -Chmdsz) ■ [N]on
^ (2)
[N}on ■ [iV2]0/[iV2]x   =   C2 ■ exp[-(k2 + kj. ■ Chmdsz)*],
where kiis the rate coefficient of the reaction N+HMDSZ, C2 is const, given by experimental conditions and Chmdsz is the HMDSZ concentration.
By dividing (2) and (1) and taking logarithm of both sides we obtain
H[N]on/[N]oS) = Co - fci ■ Chmdsz ■ t. (3)
The rate constant ki can be determined from the linear dependence (3).
For the reaction O+TEOS, the rate coefficient ki ranges from 7.1 to 9.2 [1CP15 cm3s_1] and for reaction N+HMDSZ was found k2 = (0.75 - 2.4)-KT15 [air's-1].
3  The effect of admixtures and wall recombination of atoms
The dependencies of the dissociation rates of oxygen, hydrogen and nitrogen on the admixture concentration were examined by ESR. In all experimental dependencies we can see the rapid
166
J. Janca et al.
	350 300	■	■   . ■ "	H2 . natural isotope ratio ........... \ *
	250	■ /		• •
in	200		\ •	
	150 100 50 0	m »		
0 0.5 1 1.5 2 2.5 3
Q(02), Q(H2) [seem]
Fig. 3. Ratio of nitrogen isotope abundance in afterglow as a function of the amount of O2 or H2 admixturer. Flow of nitrogen was 50 seem, corresponding to a pressure 450 Pa.
dissociation rate increase with the growing amount of admixture and after the maximum has been reached, the next increase of the amount of admixture leads to an atom concentration decrease as shown in Fig. 2. But the steepness of the increase and decrease as well as the position of the concentration maximum with respect to the admixture concentration depends on the specific gas mixture used in the experiment.
The experimental data in Fig. 2 are accompanied by results of theoretical model (for details, see [9]). It was shown that admixtures in small quantities enhance the dissociation rate due to a reduction of the wall recombination. Higher admixture concentrations cause a dissociation rate decrease due to volume reactions and admixture recombination mechanism. The rate of increase and decrease and the position of the maximum with respect to the admixture concentration differ for each gas, as the experimental conditions and dissociation and recombination properties of O2, N2 and H2 are different.
Due to different nuclear spin the 14N and 15N isotopes are easily distinguished in ESR spectra. It is interesting that adding of H2 and O2 into nitrogen gas causes also different dissociation and recombination rates of 14N and 15 isotopes (see Fig. 3).
4   Production of free electrons in afterglow
Change of atomic nitrogen concentration are for the case of oxygen admixture in nitrogen closely followed by the increase of electron density as presented in Fig. 4. The electron concentration was measured by ESR device using the absorption line of electron cyclotron resonance. Taking into account, that free electrons produced in the discharge can not reach such late positions in afterglow, we may deduce, that atomic nitrogen (most probably excited in upper electronic states) plays an important role in production of electrons in the afterglow [8].
Study of plasma processes in afterglow.
167
02/n2 [%]
Fig. 4. Dependence of electron density on O2 admixture for two positions in afterglow (distance of ESR cavity from RF discharge was 28 and 50 cm). Nitrogen flow was 50 seem and pressure 450 Pa.
Acknowledgements: This work was supported by the Ministry of Education of the Czech Republic, contract No. ME 489 and OC 527.20 and by Grant Agency of Czech Republic, contract No. 202/01/P106.
References
[1] A.A. Westenberg: Prog. React. Kinet. 7 (1973) 23
[2] N.M. Johnson, J. Walker, CM. Dolond, K. Winter, R.A. Street: Appl. Phys. Lett. 54 (1989) 1872
[3] J. Janča, A.Tálský, V. Zvoníček: Plasma Chem. And Plasma Process. 16 (1996) 187
[4] M. Janča, J. Janča, V. Kudrle, A. Tálský: in Proc. 14th Internat. Symp. Plasma Chem., Prague 1999.
(Eds. M. Hrabovský, M. Konrád, V. Kopecký), Prague 1999, p. IV-1843 [5] N.M. Johnson, J. Walker, K.S. Stevens: J. Appl. Phys. 69 (1991) 2631
[6] J. Janča, M. Janča, A. Tálský, V. Zvoníček: in Proc. 13th Internat. Symp. Plasma Chem., Peking 1997.
(Ed. C.K. Wu) 1997, p. 1-76. [7] N.M. Johnson, P.V. Santos, J. Walker, K.S. Stevens: Mater. Res. Soc. Symp. Proc. 219 (1991) 703 [8] V. Kudrle, A. Tálský, J. Janča:  in Proc. 14th Inter. Symp. Plasma Chem., Prague 1999. (Eds.
M. Hrabovský, M. Konrád, V. Kopecký), Prague 1999, p. 11-723 [9] V. Kudrle, A. Tálský, A. Kudláč, V. Křápek: Czech. J. Phys. S3 50 (2000) 305