Full text of DOI:10.1016/S0022-4073(99)00209-5

Journal of Quantitative Spectroscopy &
Radiative Transfer 67 (2000) 217}225
Spectroscopic and electric probe measurements of plasma
parameters in the negative glow region of helium discharge
plasma
A.M. Abdel-Baky*, M.H. El-Ghazaly, M.M. Mansour
Department of Physics, Faculty of Science, University of Zagazig, Egypt
Received 28 July 1998
Abstract
The essential plasma parameters, i.e., electron temperature, ¹ electron density, n and ion temperature,
%
%
¹
, have been measured in the negative glow (NG) region of a helium cold-cathode discharge plasma as
*
a function of the "ller gas pressure. For the determination of these parameters two di!erent techniques have
been used throughout this work. Spectroscopic technique depending on the line intensity ratio of He(I)
resonance line at 3888.65 As and He(II) ionized line at 4686 As , besides the width and pro"le of the working
heluim gas, is the "rst one, while electrical single probe technique depending on the voltage}current
characteristics of the plasma is the second one. By this second technique the experimental data of the #oating
potential, < , and the plasma potential, < , were used to de"ne the negative glow region. Comparison
&
ꢀ
between the results of both techniques revealed a strong agreement of the measured values of ¹ , besides
%
a fairly good agreement of the measured values of n . ( 2000 Elsevier Science Ltd. All rights reserved.
%
1
. Introduction
DC cold-cathode glow discharges have been studied for many years. Recently, e!orts have been
intensi"ed to understand the glow discharge regions, (cathode fall, negative glow and positive
column regions, respectively), which can be created from the interaction mechanism of the electron
beam with the working gas. Some of these e!orts have included experimental treatments [1}4]. The
electron beam source is produced mainly from the cathode surface and may be pulled from the
plasma sheath which is adjacent to the cathode surface. For DC-cathode plasma, the main source
of the electron beam is the primary electron released from the cathode by ion bombardment. The
*
Corresponding author.
0
022-4073/00/$- see front matter ( 2000 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 2 2 - 4 0 7 3 ( 9 9 ) 0 0 2 0 9 - 5

2
18
A.M. Abdel-Baky et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 67 (2000) 217}225
emitted electrons have been accelerated afterwards by the potential di!erence which is equal to the
cathode fall. Therefore, these electrons have gained energy to excite the gas atoms producing
the negative glow [5}7]. There is a good correlation between the length of the negative glow and
the range of accelerated electrons [8]. A study of the electron beam parameters in the normal glow
discharge can clarify the processes of the di!erent interactions and regions, the negative glow
region is one of them.
For the present investigation, optical emission spectroscopic and Langmuir probe techniques
are used to characterize the glow discharge regions which take place in the plasma. Spectroscop-
ically, the visible light emitted by the discharge plasma column is often used for diagnostic
purposes, however, due to a lack of thermal equilibrium, interpretation of the absolute intensities of
spectral lines or molecular bands is not trivial [9]. The essential plasma parameters, i.e. the electron
and ion densities, and the electron and ion temperatures, could be achieved by these techniques at
the edge and inside of negative glow region. The spectroscopic measurements depend upon the
intensity ratios, widths and pro"les of neutral and ionized He-lines. Brenning and others have
discussed the possibility of determining electron temperature from the relative intensities of
spectral lines in low-density plasmas. It is concluded that most lines can only be used at very low
densities (N (2;10ꢀꢁ m\ꢂ) because the line intensities are highly in#uenced by secondary
%
processes [10,11]. While the Langmuir probe measurements depending on the current}volt
characteristic of single electric probe have been used for measuring the plasma parameters and
de"ning the negative glow region from the experimental data of the #oating potential, < , and the
&
plasma potential, <ꢀ.
2
. The experimental system
2
.1. Source
Experiments were carried out in a #owing gas discharge tube made of Pyrex 20 cm long and
2
0 cm diameter where a pair of aluminum electrodes are sealed at its ends, Fig. 1. Helium gas was
made to #ow into one end of the discharge tube through a needle-valve until the working gas
pressure is achieved. This gas was exhausted by a rotary pump. The discharge was run from
a regulated DC-power supply with a rheostat ballast by a positive potential applying to the anode;
the cathode was earthen.
2
.2. Monochromator device and photomultiply tube
The spectroscopic light source for the present experiment is the DC glow discharge plasma
described previously. Light was collected from the discharge passed via an entrance slit of the
monochromator device. In order to obtain the necessary high resolution, 1.5 m THR1500 mono-
chromator device was used. The monochromator was "tted with a plane square, holographic
grating with 80;110 mmꢃ width-height and 1800 line/mm. The resolving power yielded while the
grating operated in a single pass is 175,000. The reciprocal linear dispersion is nearly 2.6 A/mm in
s
the wavelength range between 2000 and 9000 A. An entrance slit of 5 mm width was used. The
s
monochromator outputs were focused onto a highly sensitive photomultiplier RCA IP21 coupled
with IBM computer through ISA Divsion JOBIN-YVON spectrolink.

A.M. Abdel-Baky et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 67 (2000) 217}225
219
Fig. 1. Glow discharge plasma experiment assembly.
2
.3. Single electric probe
A single electric probe was constructed from 0.5 mm diameter tungsten wire encapsulated in
a glass sleeve. The body of the probe sites in the center of the discharge tube through the earthed
electrode. The probe could be moved along the axis of the tube by means of an external tool; the
tube wall was marked with graduations to de"ne the probe position. DC-power supply was applied
for probe potential.
3
. Experimental results and discussions
The measurements were taken at the edge and inside of negative glow plasma region at 0.4, 0.6
and 0.8 Torr. helium gas pressures for mean discharge current densities of 0.095, 0.121 and
.146 mA/cmꢃ, respectively. The measured potential di!erences between the discharge electrodes
0
were 300, 260 and 220 V, respectively, with the pressures. These measurements could be applied to
two di!erent techniques which are used to determine the electron temperature, ¹ electron density,
%
n and the ion temperature, ¹ . The "rst is spectroscopic measurement technique, and second
%
*
a single electric probe technique.
3
3
.1. Spectroscopic technique
.1.1. Ion temperature
It has been shown that the experimental spectral line pro"les have to be corrected for contribu-
tions to broadening. These contributions arise from the Doppler broadening, collisional broaden-
ing, and broadening in#uences caused by instrumental slit widths. Our measurements were
performed using a low-density glow discharge plasma. The discharge was operated in helium at
a "ler gas pressure of 0.2}0.8 Torr and at currents between 30 and 46 mA. Under these conditions

2
20
A.M. Abdel-Baky et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 67 (2000) 217}225
Table 1
Experimental He (I) and He (II) full half-widths with the calculated values of ion
temperature, ¹ , at di!erent gas pressures
*
Pressure (Torr)
Line at 3888.65 A
s
of He (I)
Line at 4686 A
s
of He (II)
*
jꢀꢄꢃ (A
s
)
¹ (eV)
*jꢀꢄꢃ (A
s
)
¹ (eV)
*
*
0
0
0
.8
.6
.4
0.382
0.376
0.374
6.49
6.29
6.22
0.448
0.453
0.418
6.150
6.288
5.354
collisional broadening are small enough to be negligible [12,13]. The instrumental pro"le of the
monochromator was determined experimentally by investigating the width of the 6328 A emission
line of He}Ne laser. The laser line had a full-width at half-maximum intensity (FWHM) of less than
.05 A. So it is easy to conclude that, the Doppler broadening makes the largest contribution to the
s
0
s
experimental line width.
As we knew the instrumental width, it was possible to deduce the ion temperature from Doppler
line broadening [14,15]. Since the emitting atoms and ions in the plasma possess random velocities,
spectral lines will be broadened due to the Doppler e!ect. If the random motion is of thermal
origin, the Doppler pro"les are Gaussian (in the absence of any other broadening mechanism) with
half-width (at half-maximum intensity) given by [16]
2
C
k¹ lꢃ ꢀꢄꢃ
G ꢅ
*
l""
,
(1)
Mcꢃ
D
where ¹ is the temperature of the emitting ions and M is their mass.
*
The experimentally determined full half-widths of He (I) resonance line at 3888.65 A
line at 4686 A
Table 1.
s
and He (II)
s
, with the calculated values of ion temperature at di!erent gas pressures, are given in
3
.1.2. Electron temperature and denisty
Many of the spectroscopic diagnostic methods of measuring electron temperature and density
are based on the theoretical plasma models [17]. It should be mentioned that, for the experimental
conditions here reported, the electron density inferred from the probe measurements gives
n +10ꢀꢅ cm\ꢂ. This suggested that the so-called steady-state corona model should be used in
%
studying the plasma formed in the glow discharge.
The corona mode, i.e. the collisional ionization (or excitation) is balanced by radiative recombi-
nation (or spontaneous decay), was developed for plasmas of low electron density. For hydrogen
and hydrogen-like ions, the electron density may be expressed using the following inequality [17]:
1
C
.162;10ꢂ(z#1)ꢃ
n (5.6;10ꢆ(z#1)ꢁ¹ꢀꢄꢃexp
,
D
(2)
%
%
¹%
where ¹ is the electron temperature.
%

A.M. Abdel-Baky et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 67 (2000) 217}225
221
Table 2
The electron temperature, ¹ and density, n values for nega-
%
%
tive glow region of He glow discharge as a function of the He
pressure
Pressure (Torr)
¹ (eV)
n%(10ꢀꢂ cm\ꢂ
%
0
0
0
.8
.6
.4
5.723
5.839
6.287
3.962
3.997
4.127
Accurate spectroscopic measurements of ¹ , under the steady-state corona model, could be
%
obtained from the ratio of ion and neutral line intensities. In the present study, ¹ can be obtained
%
from the ratio of the ion and neutral line intensities (i.e. 4686 He (II)/3888.65 A
the following relation [14]
s
He (I) ratio) using
Iꢀ f ꢀgꢀjꢂ
E !Eꢀ!E #E S
ꢇ ꢇ
"
exp
,
(3)
I
fgjꢀꢂ
C
k¹%
D
a
where I, j, g, and f are total intensity, wavelength, statistical weight (of the lower state of the line),
and absorption oscillator strength, respectively, of the neutral line and E, E are its excitation and
ꢇ
ionization energy. The primed quantities refer to the ionized line. S and a are collisional ionization
and radiative recombination coe$cients given by [17]
f¹ꢀꢄꢉ
s(z, g)
k¹%
S(¹ , z, g)"2.34;10\ꢈ
%
exp !
(4)
(5)
%
[m(z, g)]ꢈꢄꢉ
C
D
and
a(¹ , z, g)"2.05;10\ꢀꢃ
s(z!1, g),
%
¹ꢀꢄꢃ
%
where m is the number of outer electrons, ¹ is in K, and s(z, g) is the ionization potential of the ion
%
of charge z in its ground level [17].
The results of ¹ obtained using Eqs. (3)}(5) as a function of the gas pressure in helium glow
%
discharge in the negative glow region are given in Table 2. Values of ¹ were varied from nearly
%
5
}7 eV in the pressure range of 0.4}0.8 Torr. The present measurements show that, ¹ values
%
decrease as the gas pressure increase. This can be explained as, at low gas pressure, the elec-
tron}neutral particles collision frequency is small and the mean free path is high. While at high
pressures the collision frequency increases and hence, the loss of electron energy is high, thus
¹%
decreases [18].
The electron density, n may be estimated using inequality (2) with the help of the values of
which was determined previously. Table 2 includes the values of electron density, n as
%
¹
%
%
a function of gas pressure. The present results indicate that the electron density increases by
increasing the gas pressure. This may be attributed to the increase in the number of electron}atom
collisions, and this consequently increases the ionization rate [18].

2
22
A.M. Abdel-Baky et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 67 (2000) 217}225
3
.2. Electrical single-probe technique
The used probe made of tungsten wire 0.5 mm radius gave the I}< characteristic curves. The
analysis of these curves showed that the electron temperature could be derived from Eq. (6) [19]
ln i "(e/k¹ )< #ln A j !(e/k¹ )< ,
(6)
%
%
ꢁ
ꢁ ꢅ
where < is the probe potential, < is the space potential, A is the e!ective area of probe-surface-
%
ꢀ
ꢁ
ꢀ
ꢁ
exposed plasma, i is the electron probe current and j is the random electron current density. By
%
ꢅ
plotting ln i against the probe potential, < one can easily estimate the electron temperature, ¹ .
%
ꢁ
%
The plasma electron density was determined from the saturation electron probe current at the
space potential by Eq. (7) [20]
i "i exp(g) (g40)
%ꢀ
(7)
%
and g"e(< !< )/k¹ . When the probe potential, < , approaches to the space potential, < , the
ꢁ
ꢀ
%
ꢁ
ꢀ
saturation electron current becomes
i "A e n (k¹ /2p m )ꢀꢄꢃ (g"0).
(8)
%ꢀ
ꢁ
%
%
%
All probe measurements were carried out on the electrons when the probe was located at
di!erent axial positions inside 20 mm length of the plasma, from the cathode surface to the negative
glow (NG) region. Considering that there is a weak plasma in equilibrium state and it covers the
Fig. 2. Floating potential vs. distance from the cathode surface to inside the negative glow region.

A.M. Abdel-Baky et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 67 (2000) 217}225
223
Fig. 3. Plasma potential vs. distance from the cathode surface to inside negative glow region.
Fig. 4. Plasma temperature vs. distance from the cathode surface to inside negative glow region.

2
24
A.M. Abdel-Baky et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 67 (2000) 217}225
Fig. 5. Electron density vs. distance from the cathode surface to inside negative glow region.
mentioned length, the #oating potential and plasma spatial potential pro"les can be deduced from
the single probe characteristics. The #oating potential, < , is the probe potential at which the
&
electron and ion currents are equal, while the plasma potential (space potential), < , is the potential
ꢀ
of the point of interaction between the electron saturation current line and the line of the increasing
electron current of the ln (I)}< curve.
By both #oating and plasma potential pro"les we were able to identify plasma regions inside the
length of 20 mm from the cathode surface, the negative glow, NG, region is one of them. Figs. 2
and 3 and show that the NG region takes places in the discharge tube at distances of 5, 7 and
1
0 mm away from the cathode surface at pressures of 0.8, 0.6 and 0.4 torr, respectively. It has
also been noted that both the #oating and plasma potential points are nearly convergent inside
the lighted regions, such as the negative glow, NG, region. This is because of the accumulation of
the majority electrons inside these regions. The collected electrons are subjected to electron}atom
collision (inelastic collision) and therefore loosing most of their energies.
For optimum conditions of DC-glow discharge, variation of the electron temperature, ¹ and
%
density, n to the plasma column obtained, is shown in Figs. 4 and 5. Electron temperature, ¹ was
%
%
varied in the range of 4.0}8.0 eV at pressure values of 0.8}0.4 Torr. Meanwhile, the electron density,
n% takes the values of 0.9}3.5;10ꢀꢅ cm\ꢂ for the mentioned pressures.

A.M. Abdel-Baky et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 67 (2000) 217}225
225
4
. Conclusion
An important point of view is the comparison between the results of the measured electron
temperature, ¹ and electron density, n from the two techniques used. It is a point of interest to
%
%
conclude that, a strong agreement exists between the measured values of, ¹ while there is a fair
%
agreement of the measured values of, n . The fair agreement of, n is because of the shortcoming of
%
%
the spectroscopic methods for determining the electron density, n .
%
References
[
[
[
[
[
[
[
1] Doughty DA, Den Hartog EA, Lawler JE. Phys Rev Lett 1987;58:2668.
2] Doughty DA, Lawler JE. J Appl Phys 1984;45:611.
3] Shoemaker JR, Ganguly BN, Garscadden A. Appl Phys Lett 1988;52:2019.
4] Den Hartog EA, Doughty DA, Lawler JE. Phys Rev A 1988;38:2471.
5] Cobine JD. Gaseous conductors. New York: Dover, 1958.
6] Marti A, Abril L, Valles JA, Abarca. Thin Solid Films 1985;124:59.
7] Mase H, Unuma M, Tanabe T, Ikehata T. Proceedings of the Eighth Symposium on Dry Process (IEE) PN, Tokyo,
1
986, p. 18.
[
[
8] Sanborn CB, Introduction to electrical discharges in gases. New York: Wiley, 1966.
9] Behringer K, Fantz U. J Phys D 1994;27:2128.
[
[
[
[
[
[
[
[
[
[
[
10] Schmitt PS. Inst Tech 1975;22:35.
11] Brenning N. J Phys D 1980;13:1459.
12] Wagenear HC, Pickford J, DE Galan L. Spectrochim Acta 1974;29B:211.
13] Hannaford B, McDonald DC. J Phys B 1978;11(7):1177.
14] Greim HR, Plasma spectroscopy. NewYork: McGraw-Hill, 1964.
15] Platisa M, Dimitrijevic M, Popovic M, Konjevic N. J Phys B 1977;10(15):2997.
16] DonoHue RG, Majkowski RF. J Appl Phys 1962;33(1):3.
17] Huddlestone RH, Leonard SL. Plasma diagnostic techniques. New York: Academic Press, 1965.
18] Attya MA. Ph.D. Thesis. Zagazig University, 1996.
19] Johnson EO, Malter L. Phys Rev 1950;80(1):58.
20] Mantel TD. Proc SPIE Int Soc Opt Eng(USA) 1991;1392:466.