Study on the effect of target on plasma parameters of magnetron sputtering discharge

plasma
P. Saikia, B. Kakati, and B. K. Saikia

Citation: Physics of Plasmas 20, 103505 (2013); doi: 10.1063/1.4825235

View online: http://dx.doi.org/10.1063/1.4825235
View Table of Contents: http://scitation.aip.org/content/aip/journal/pop/20/10?ver=pdfcov
Published by the AIP Publishing

Articles you may be interested in

Study on the effect of hydrogen addition on the variation of plasma parameters of argon-oxygen magnetron glow
discharge for synthesis of TiO2 films
AIP Advances 6, 045206 (2016); 10.1063/1.4947091

Direct measurement and modeling of the redirected ion flux in a high-powered pulsed-plasma magnetron
J. Vac. Sci. Technol. A 33, 031301 (2015); 10.1116/1.4914174

Plasma diagnostics of low pressure high power impulse magnetron sputtering assisted by electron cyclotron
wave resonance plasma
J. Appl. Phys. 112, 093305 (2012); 10.1063/1.4764102

Drifting localization of ionization runaway: Unraveling the nature of anomalous transport in high power impulse
magnetron sputtering
J. Appl. Phys. 111, 053304 (2012); 10.1063/1.3692978

Dynamics of reactive high-power impulse magnetron sputtering discharge studied by time- and space-resolved
optical emission spectroscopy and fast imaging
J. Appl. Phys. 107, 043305 (2010); 10.1063/1.3305319

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 152.88.96.146 On: Sat, 03 Sep
2016 11:39:27
 PHYSICS OF PLASMAS 20, 103505 (2013)

Study on the effect of target on plasma parameters of magnetron sputtering

discharge plasma
P. Saikia,a) B. Kakati, and B. K. Saikia
Centre of Plasma Physics, Institute for Plasma Research, Nazirakhat, Sonapur-782 402, Kamrup, Assam, India
(Received 16 April 2013; accepted 1 October 2013; published online 16 October 2013)
In this study, the effect of magnetron target on different plasma parameters of Argon/Hydrogen
(Ar - H2) direct current (DC) magnetron discharge is examined. Here, Copper (Cu) and Chromium
(Cr) are used as magnetron targets. The value of plasma parameters such as electron temperature
(kTe), electron density (Ne), ion density (Ni), degree of ionization of Ar, and degree of dissociation
of H2 for both the target are studied as a function of input power and hydrogen content in the
discharge. The plasma parameters are determined by using Langmuir probe and Optical emission
spectroscopy. On the basis of the different reactions in the gas phase, the variation of plasma
parameters and sputtering rate are explained. The obtained results show that electron and ion
density decline with gradual addition of Hydrogen in the discharge and increase with rising input
power. It brings significant changes on the degree of ionization of Ar and dissociation of H2.
The enhanced value of electron density (Ne), ion density (Ni), degree of Ionization of Ar, and
degree of dissociation of H2 for Cr compared to Cu target is explained on the basis of it’s higher
Ion Induced Secondary Electron Emission Coefficient (ISEE) value. V C 2013 AIP Publishing LLC.

[http://dx.doi.org/10.1063/1.4825235]

I. INTRODUCTION few reports are available about the dependence of plasma pa-
rameters of magnetron discharge on the target conditions
Magnetron sputtering has been developed rapidly as an
namely; the target material and degree of chemisorptions in
established tool of wide range of industrially important coat-
case of reactive sputtering.9 The target conditions mostly
ings over the last decade.1 In the basic sputtering process, a
depend on its ISEE coefficient values. The target current-
target (or cathode) is bombarded by energetic ions generated
voltage (I-V) characteristics of magnetron sputtering are also
in glow discharge plasma situated in front of the target. This
determined by ISEE coefficient9 which in turn can influence
process can lead, during prolonged ion or plasma bombard-
the plasma parameters of magnetron discharge. Ion flux
ment of a material, to significant erosions of material.
flowing to the substrate is a function of target current. As the
Secondary electrons are also generated from the target sur-
ion flux is crucial for the properties of deposited film, thus
face as a result of the ion bombardment. However, this pro-
ISEE coefficient values can affect it significantly. Several
cess is limited by low deposition rate, substrate heating
authors have published empirical relations for the ISEE coef-
effect, and lower ionization coefficients.2 In magnetron sput-
ficient as10–12
tering device, a magnetic field parallel to the cathode surface
is used to constrain the secondary electron motion to the vi- cISEE ¼ 0:032ð0:78Ei  2/Þ; (1)
cinity of the target.3 It does increase the probability of ioniz-
ing electron atom collision and hence dense plasma near the cISEE ¼ 0:16ðEi  2/Þ; (2)
target region. This in turn leads to increased ion bombard-
cISEE ¼ 0:2ð0:8Ei  2/Þ=EF : (3)
ment of the target giving higher deposition rate at the sub-
strate. Thus, secondary electrons emitted from the target
Ei the ionization energy of the ion, u the work function
plays an important role in magnetron sputtering device.
and EF the Fermi energy of the metal. Based on these empiri-
Amount of secondary electrons emitted per ion is known as
cal relations, an averaged value of 0.091 for Chromium (Cr)
ion induced secondary electron coefficient (ISEE). Phels
and 0.082 for Cupper (Cu) target materials were calculated.
et al.,4 have shown that ISEE coefficient strongly depends on
In our earlier study, the effect of SEE on dust charging is
the condition of target material and independent of bombard-
studied in low pressure filament discharge.13 In this paper, the
ing ion types for ion energy below 500 eV.
experimental measurement of influence of target material on
Most of the works in the literature using magnetron
plasma parameters of magnetron discharge is reported. Cu
sputtering have commonly reported the properties of depos-
and Cr targets are selected because of their wide range of util-
ited films.5–7 Many experimental and theoretical works have
ity as coating materials and distinct ISEE coefficient values.
been performed on the physical properties8 of magnetron dis-
Cr and Cu are sputtered deposited on Si(100) substrate in
charge and its dependence on process parameters like gas
Argon/Hydrogen (Ar/H2) plasma at different discharge condi-
pressure, substrate biasing, electrical power etc. But very
tions. First, working pressure is varied by adding more hydro-
gen at a fixed Argon flow rate in suitable pressure range.
a)
Electronic mail: partha.008@gmail.com. These types of experiments are interesting to understand the

1070-664X/2013/20(10)/103505/9/$30.00 20, 103505-1 C 2013 AIP Publishing LLC

V

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 152.88.96.146 On: Sat, 03 Sep
2016 11:39:27
 103505-2 Saikia, Kakati, and Saikia
TABLE I. Discharge voltage and current at various working pressure for Chromium and Copper target.
Target
Working pressure (101 Pa) Voltage (V)
4 587
6 598
8 605
10 615
variation of plasma parameters of discharge during reactive
sputtering.
Second, at a fixed working pressure, these parameters
are studied as function of applied power. I-V characteristics
of planar magnetron at various deposition conditions are
determined. Finally, measured values of ion density, electron
density, degree of ionization and sputtering rate as a function
of working pressure and applied power for both targets are
presented in this investigation.
II. EXPERIMENTAL SETUP
The films were sputter deposited in a stainless steel
plasma chamber with a diameter of 45 cm and length 50 cm
on Si wafers. The magnetic field strength at the centre of the
cathode reached 600G. The water cooled Cu and Cr target of
diameter 70 mm were made as cathode of the planar magne-
tron sputtering system. During the deposition Argon (Ar)
and Hydrogen (H2) are feed into the deposition chamber by
two digital mass flow controllers (AALBORG made).
Working pressure in the chamber is varied by increasing
hydrogen concentration [H2] in the argon discharge from 0
to 18.9% at a constant argon flow rate (8.5 SCCM). The con-
centration of H2 in feed gas is calculated comparing the par-
tial pressure of both the gases. For different percentage of
hydrogen flow rate, the total working pressure changes from
4  101 Pa to 10  101 Pa. Again, at a constant working
pressure (4  101 Pa) keeping the other conditions same,
effective load power is varied from 50 W to 300 W. The
details of the currents and voltages at different discharge
conditions for both Cr and Cu targets are given in Tables I
and II. The chamber is evacuated by a diffusion pump
backed by a rotary pump up to base pressure of 104 Pa. A
single cylindrical probe with a length of 3.0 mm and a diam-
eter of 0.1 mm is installed in the discharge chamber. The
probe measurement is carried out at the discharge center in
the downstream region of 70 mm from the cathode, i.e.,
10 mm from the substrate. Plasma parameters such as
TABLE II. Discharge voltage and current at various applied power for chro-
mium and copper target.
Target Chromium Copper
Input power (W) Voltage (V) Current(A) Voltage (V) Current (A)
50 452 0.11 510
100 500 0.21 550
200 562 0.36 606
300 600 0.51 650
Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 152.88.96.146 On: Sat, 03 Sep
Phys. Plasmas 20, 103505 (2013)
Chromium Copper
Current (A) Voltage (V) Current (A)
0.52 627 0.481
0.513 635 0.472
0.507 645 0.465
0.493 668 0.451
electron temperature (Te in eV), electron density (Ne), and
ion density (Ni) are measured with the probe in practically
magnetic field free region.
To estimate the relative contribution of hydrogen ions and
argon ions in the plasma due to the mixing of different percent-
age of hydrogen gas in argon plasma through measuring the in-
tensity level of specific emission, a 1=2 m digikrom spectrometer
(CVI Laser Corp, USA. Digikrom Model DK 480) is used. The
system consists of a photo multiplier tube (PMT: Model
AD110, wavelength range: 185–930 nm) and a grating with
1200 grooves/mm for detection in the region k ¼ (400–850) nm.
The deposition rate at various discharge conditions is monitored
by Crystal Thickness Monitor (DTM-101) made by
HIND-HIVAC, Bangalore. The monitor head was placed at the
same horizontal plan as with the substrate holder.
III. RESULTS AND DISCUSSION
A. Influence of target material on I-V characteristics of
magnetron discharge
The I-V characteristics of magnetron discharge has been
proposed by Westwood et al.:14
I ¼ bðV  V0 Þ2 ; (4)
with V0 is minimum voltage required to maintain the dis-
charge, b is a constant which is a measure of steepness of
I-V characteristics. Value of b can be determined by a linear
fit to square root of current (I) as a function of voltage. It has
been shown by Deepla et al.9 that b increases with increasing
ISEE coefficientpffiffi values. Now, at fixed working pressure of
4  101 Pa, I  V characteristics of both targets are plot-
ted in Fig. 1(a) and b (105 A/V2) values were found to be
34.1 for Cr and 21.9 for Cu. Higher value of b for Cr than
Cu justifies the theoretically calculated values of ISEE coef-
ficient for the targets.
To study the influence of working pressure on discharge
properties at fixed target current, discharge voltage is plotted
as a function of working pressure in Fig. 1(b). As expected, we
notice a abrupt decrease of the discharge voltage with increas-
ing working pressure up to 3.0  10 1 Pa. After that, with fur-
ther increase of working pressure leads to only a slight change
of discharge voltage. Initial decrease of discharge voltage is
due to strong probability of recapture of electrons emitted
from the cathode surface at low working pressure.15 Increase
0.10
0.19
of working pressure results a higher secondary electrons yield
0.33 and lowering of discharge voltage. Now, magnetron I-V char-
0.47 acteristics were determined at different working pressures
(0.2 Pa, 0.4 Pa, 0.6 Pa, 0.8 Pa, and 1.0 Pa) and corresponding b
2016 11:39:27
 103505-3 Saikia, Kakati, and Saikia
pffiffi
FIG. 1. (a) I  V characteristics for Chromium and Copper target at working pressure 4  101 Pa. (b) Discharge voltage as a function of the working pres-
sure for Chromium and Copper target at fixed discharge current.
values were evaluated. It is plotted as a function of working
pressure in Fig. 2. As expected, increasing working pressure
lowers the value of b and after 4  101 Pa the change is in-
significant. As, the present investigation of the physical proper-
ties of discharge is done in the working pressure range from
0.4 Pa to 1.0 Pa, it can be said that the target current and hence
ion current flowing to the substrate are almost constant in this
range and it is true for both the target materials.
B. Description of ionic processes in H2 containing
Argon plasma
Study of basic gas phases is essential to understand the
effect of H2 addition to pure Ar plasma. When H2 is added to
FIG. 2. Influence of working pressure on the value of b.
Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 152.88.96.146 On: Sat, 03 Sep
2016 11:39:27
Phys. Plasmas 20, 103505 (2013)
Argon, besides Ar atoms (in ground state and excited to met-
astable states (Arm*) at energy 11.2 eV), Arþ ions and elec-
trons, several hydrogen species including Hþ, H2þ, H3þ
ions; ArHþ ions, the ground state H atom, ground state H2
molecule and various electronically excited states of H2 mol-
ecules are also possible.16 All those species in plasma
undergo different chemical reaction with different the rate
constants and relative cross sections. The reactions which
are essential for the domain of this work are mentioned in
Table III. With increasing hydrogen addition a drop in elec-
tron and Arþ ion densities are experimentally observed in
the literature.17–20 On the other hand, the densities of hydro-
gen related ions, i.e., Hþ, H2þ, H3þ, and ArHþ are expected
to increase with H2 addition for obvious reasons. The effect
is most pronounced for Hþ, H2þ, H3þ and is slightly less sig-
nificant for ArHþ ions.
Addition of Hydrogen in Argon plasma causes Arþ ion
loss predominantly through the charge transfer (Reaction 1
in Table III) or H atom transfer (Reaction 2 and Reaction 3)
forming different hydrogenic species (ArHþ, H, H2þ)16
Ar þ þ H2 ! ArðfastÞ þ H2 þ ; (5)
Ar þ þ H2 ! ArHþ þ H; (6)
Ar þ H2 þ ! ArHþ þ H: (7)
The major loss mechanism for electrons in such plasma
is electrons recombination with ArHþ (Reaction 4), H2þ
(Reaction 5), and H3þ (Reaction 6)
e þ ArH þ ! Ar þ H; (8)
e þ H2 þ ! H þ H; (9)
e þ H3 þ ! H þ H þ H: (10)
 103505-4 Saikia, Kakati, and Saikia Phys. Plasmas 20, 103505 (2013)

TABLE III. Arþ, ArHþ, Arm* and Hydrogen like ion production and loss processes.

No. Reaction rate Constant (K) Name

þ þ 11 3 1
1. Ar þ H2 ! Ar(fast) þ H2 8.0  10 cm s Charge transfer23
2. Arþ þ H2 ! ArHþ þ H 6.0  1010 cm3s1 H atom transfer23
3. Ar þ H2þ ! ArHþ þ H 1.7  109 cm3s1 Proton transfer23
4. e þ ArHþ ! Ar þ H 1.7  107 cm3s1 Recombination23
5. e þ H2þ ! H þ H 1.0  107 cm3s1 Recombination16
6. 
e þ H3þ ! H þ H þ H 1.0  107 cm3s1 Recombination16
7. ArHþ þ H2 ! Ar(fast) þ H3þ 1.5  109 cm3s1 Proton transfer23
7. H2þ þ Ar ! Arþ þ fast H2 2.2  1010 cm3s1 Charge transfer23
8. H2 þ Hþ ! H(fast) þ H2þ 2  109 cm3s1 Charge transfer16
9. ArHþ þ Ar ! Ar(fast) þ Hþ þ Ar 1.5  109 cm3s1 Induced ionization16
10. Hþ þ Ar ! H(fast) þ Arþ 2.2  1010 cm3s1 Charge transfer16
11. Ar þ e ! Arm* þ e 1.0  011 cm3s1 Ionization16
12. Arm* þ H2 ! Ar þ H þ H 7  1011 cm3s1 Quenching dissociation16

Reaction 2 and Reaction 3 are responsible for formation 0:61eNi uB ¼ Ji ; (15)

of ArHþ ions. In such plasma, the densities of H3þ ions are
reported to be several orders of magnitude higher21,22 than
H2þ and Hþ densities in the literature. H3þ is mostly formed
through proton transfer (Reaction 7 and Reaction 8).
ArH þ þ H2 ! ArðfastÞ þ H3 þ ;
H2 þ þ H2 ! H þ H3 þ :
The Ar metastable atoms are mainly created by electron
impact excitation (Reaction 9), followed by fast Ar0 impact
excitation and fast Arþ impact excitation and relative impor-
tance of these processes are more or less same, while
quenching upon collision with H2 (Reaction 10) are main
loss mechanism16 for Arm* atoms.
Ar þ e ! Arm  þ e
Arm  þ H2 ! Ar þ H þ H:
Also, it is evident from Reaction 1 and Reaction 7 that
production of Ar (fast) atoms decline with the reduction of
Arþ and ArHþ ions.
C. Plasma parameters as a function of working
pressure and input power
1. Variation of secondary electron density in the ring
region of the magnetron discharge
There are two categories of electrons in DC planar mag-
netron discharge. Fast secondary electrons (hot electrons
emitted at the cathode) and bulk electrons created in the
main discharge region. Those electrons that have enough
energy to ionize neutrals include mostly fast secondary elec-
trons. In the ring region23 of the magnetron discharge, the
hot secondary electrons are responsible for ionization of neu-
trals and thus eventually generating electrons in the plasma
bulk. Estimating the value of plasma density within the ring
region and using the data for the ionization cross section for
secondary electrons, it is possible to determine the value of
hot secondary electron density (Ns) within the ring region.23
The Bohm flux can be used to estimate Ni in the ring
region of the discharge and can be given as
where the electron temperature Te enters only weakly. Here,
Ji is ion current density and uB is the Bohm velocity of ions.
Now, ion density (Ni) can be related to fast secondary elec-
tron density (Ns) within this region as
(11)
Ni ¼ Na Ns ri v; (16)
(12)
where v is the drift velocity of the electrons, Na is neutral
density, and ri is the ionization cross section for secondary
electrons ( 5  1017 cm2). Using above mentioned equa-
tions, the value of fast secondary electron density within
the ring region is evaluated. It is observed that with increas-
ing working pressure the secondary electron density
decreases from 7.8  1010/m3 to 2.67  1010/m3 for Cr tar-
(13) get and 6.67  1010/m3 to 2.26  1010/m3 for Cu target.
While an opposite trend of rising secondary electron den-
(14) sity from 3.34  1010/m3 to 7.5  1010/m3 for Cr target and
2.56  1010/m3 to 6.67  1010/m3 for Cu target as a func-
tion of increasing applied power (50 W to 300 W) are also
found. As expected the value of ion density (of the order of
1017/ m3) is found to be much higher than hot secondary
electron density (of the order of 1010/m3) in this region of
the discharge. Also, it is interesting to mention that density
of hot secondary electron decreases when hydrogen is
added to the pure argon discharge. Using Langmuir probe
diagnostics, the density of bulk electrons are measured in
the substrate vicinity and results will be presented in the
subsequent sections. It is observed that the density modula-
tion of bulk electrons as a function of deposition parameters
is similar to hot secondary electron density modulation.
2. Variation of electron temperature (Te) and electron
density (Ne) in the substrate vicinity of magnetron
discharge
A single Langmuir probe is used to deduce the electron
temperature and electron density of plasma discharge. The
evolution of electron density (Ne) and the ion saturation
current density (Ii (V)) with [H2] and input power are
shown in Figs. 3 and 4. It can be seen that hydrogen
addition to Argon plasma drastically reduce Ne from 1.4

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 152.88.96.146 On: Sat, 03 Sep
2016 11:39:27
 103505-5 Saikia, Kakati, and Saikia Phys. Plasmas 20, 103505 (2013)

processes. As a result Ne drops as a function of working

FIG. 3. Variation of electron density (Ne) and ion saturation current (Ii (V))
as a function of working pressure for Cr and Cu target.
pressure. The increase in input power increases the colli-
sion between electrons and neutrals which in turn enhances
the electron density as well as ion saturation current density
(Ii (V)) gradually for both Cr and Cu, respectively.
For such kind of plasma, a prominent change in electron
temperature is also observed. With increasing working pres-
sure, value of Te increases significantly from 3.3 eV to
6.60 eV for Cr and from 5.20 eV to 7.8 eV for Cu. Since in
the ohmic heating regime there is an inverse relation
between Ne and Te, the observed gradual increase of Te upon
hydrogen addition and decline of Te as a function of input
power can be easily understood [Fig. 5]. The higher value of
bulk electron density in the magnetron discharge with Cr tar-
get than Cu target is due to its high cISEE coefficient value
(0.091) compared to Cu (0.082). By particle simulation
Kondo et al.,8 have shown that the plasma density in the
bulk region of the plasma is larger for higher value of cISEE .
As the electron density value is higher for Cr compared to
Cu one could expect lower value of electron temperature
(kBTe) for it.

 1015m3 to 6.46  1014 m3 for Cr and 9.5  1014 m3

to 5.8  1014 m3 for Cu target, respectively. As the elec-
tron density decreases as a function of working pressure,
the ion density should follow the similar behavior if one
refers to the quasi neutrality property of plasma. With this
consideration, it is found that ion saturation current density
(Ii (V)) declines with hydrogen enrichment in argon plasma.
The decreasing value of electron density reflects major
electron loss rate when H2 is added to Argon plasma. In
Ar-H2 plasma the dissociative recombination reactions
between electrons and molecular ions (Hþ, H2þ, and H3þ)
formed in the plasma tend to more pronounce than ambipo-
lar diffusion and Arþ- electron recombination.19,24 Due to
presence of these additional loss processes with hydrogen
introduction in such plasma, the electron consumption rate
becomes more significant than various electron production

FIG. 4. Influence of input power on electron density (Ne) and ion saturation FIG. 5. Influence of (a) working pressure and (b) input power on variation
current (Ii (V)) for Cr and Cu target. of electron temperature (kTe) for Cr and Cu target.

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 152.88.96.146 On: Sat, 03 Sep
2016 11:39:27
 103505-6 Saikia, Kakati, and Saikia Phys. Plasmas 20, 103505 (2013)

3. Variation of mean ion mass (Mi) and ion density (Ni)

As Ar/H2 plasma contains various ion species (Arþ,
ArH , Hþ, H2þ, H3þ etc.) its essential to define ion density
þ

with respect to mean ion mass (Mi). Also, mean ion mass can
provide compensation effect of the poor dependence of ion
saturation current density value upon [H2] addition. Using the
ion saturation current (Ii (V)), electron temperature (kTe) and
ion density (Ni) values, the mean ion mass can be calculated25
and is given as a function of working pressure for both the tar-
gets in Fig. 6. It is clear from the figure that presence of
hydrogenic ion decreases the mean ion mass from 40 a.m.u to
25.11 a.m.u. for Cr target and 40 a.m.u to 26.42 a.m.u for Cu
target, respectively. From the above data and using the fact
that total ion density (Ni) is the sum of Ar-like (Arþ, ArHþ)
and hydrogenic ions (Hþ, H2þ, H3þ), Ar like ion density is
calculated [Fig. 6]. Ar like ion density decreases linearly with
hydrogen percentage enrichment in discharge. Addition of
hydrogen in Ar plasma reduces Arþ ion predominantly
through charge transfer between Arþ and H2 (Reaction 1) or
H atom transfer between the same (Reaction 2). This trend is
found to be similar for both types of targets.
The mean ion mass and density of Ar like ions in the
discharge as a function of input power is shown in Fig. 7.

FIG. 7. Variation of mean ion mass and Ar like ion density as a function of
input power for Cr and Cu target.

With increasing input power mean ion mass rises and a

corresponding increase in density of Ar like ions from 2.16
 1014m3 at 50 W to 5.36  1014m3 at 300 W for Cu target
and 2.26  1014 m3 at 50 W to 7.39  1014 at 300 W for Cr
target are observed. It is due to the fact that increase in input
power increases the degree of ionization as explained in the
previous section. The observed enhanced value of Ar like ion
density for Cr target than Cu target in the discharge is due to
higher ISEE coefficient of Cr than Cu.

4. Degree of argon ionization and H2 dissociation from

FIG. 6. Variation of mean ion mass and ar like ion density as a function of
working pressure for Cr and Cu target.
OES study
The intensity of spectral emission line is closely related
to density of the species in ground state and hence is often
used to deduce the evaluation of density.26 Thus, the line in-
tensity ratio between two suitable spectral lines can be used
to estimate the value of degree of ionization and degree of
H2 dissociation. Fig. 8 shows the variation of Ar I line
(750.38 nm), Ar II line (476.48 nm), Ha line (656.28 nm),
Ar(I) line (811.53 nm) intensity as a function of working
pressure for Cr and Cu target respectively. Here, the intensity
ratio of Argon spectral line at 476.4 nm (Ar II) to at 750.38
(Ar I) (d) is used to estimate the value of degree of ioniza-
tion. These two lines are selected as the excitation thresholds
are very close and the transition probabilities (6.4  107 s1
at 476.5 nm and 4.0  107 s1 at 751.5 nm, respectively)

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 152.88.96.146 On: Sat, 03 Sep
2016 11:39:27
 103505-7 Saikia, Kakati, and Saikia Phys. Plasmas 20, 103505 (2013)

FIG. 9. Variation of the intensity ratio (d) of Argon spectral line at 476.4 nm
(Ar II) to at 750.38 (Ar I) as a function of (a) working pressure (b) input
power for Cr and Cu target.

FIG. 8. Observed Intensity (I) of atomic argon line at 476.48 nm (Ar II),
750.38 nm (Ar I), 811.53 nm (Ar I) and Ha line (656.28 nm) as a function of
working pressure (a) Cr target (b) Cu target.

are almost similar27,28 for them. Fig. 9(a) shows that with
increase of working pressure, d decreases from 7.1  102
to 5.8  102 for Cr and 4.3  102 to 3.6  102 for Cu
target and with increase of input power an opposite trend is
observed [Fig. 9(b)]. From Langmuir Probe study, we have
already seen that density of Argon like ions drop as a func-
tion of H2 addition and increases as a function of input
power. It explains corresponding increase and decline in the
value of d. It is reasonable to find higher value of d for Cr to
Cu irrespective of discharge condition. As electron impact
ionization of Ar atom is major production mechanism for
Arþ ions, higher electron density for Cr will naturally lead to
enhanced value for d compared to Cu.
In this study, the ratio of Ha line (656.28 nm) to Ar(I) line
(811.53 nm) (c) can be used to determine the value of degree
of H2 dissociation (fD).29,30 The variation of c as a function
of working pressure and input power is given in Figs. 10(a) FIG. 10. Variation of intensity (I) ratio of Ha line (656.28 nm) to Ar(I) line
and 10(b), respectively. It can be seen that c decreases from (811.53 nm) as a function of (a) working pressure (b) input power for Cr and
7.2% to 2.94% for Cr and 2.8% to 2.35% for Cu target with Cu target.

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 152.88.96.146 On: Sat, 03 Sep
2016 11:39:27
 103505-8 Saikia, Kakati, and Saikia
hydrogen enrichment in the discharge. In Ar/H2 plasma, the
dominant production of H atoms is due to dissociative excita-
tion of H2 molecules by Arm* atoms.16 With addition of H2,
number of Arm* atoms decrease continuously. As the quench-
ing dissociation (due to reaction 12) is the most important
mechanism for formation of H atoms, it indeed explains the
drop in c. On the other hand, increase of c as a function of
input power [Fig. 10(b)] can be related to the fact that produc-
tion rate of Arm* atoms increases with increase of input
power. The production processes of Arm* involves impact ex-
citation due to electrons, Arþ ions and fast Ar atoms. As elec-
tron density and Ar like ion density increase gradually in this
case, corresponding increase in c is observed. Here also, we
have observed a relatively higher value of c for Cr target com-
pared to Cu at most of the discharge conditions.
D. Hydrogen effect on the deposition rate of target
In order to correlate the observed plasma properties with
physical nature of sputtering process, we have measured the
deposition rate for various discharge conditions. The result is
given in Figs. 11(a) and 11(b), respectively. It is seen
that the sputtering rate is maximum with pure Argon
plasma compared to hydrogen additive Ar plasma. Addition
of Hydrogen gradually decrease the sputtering rate from
2.6 A0 s1 to 2.1 A0 s1 for Cr and 5.6 A0 s1 to 4.3 A0 s1
for Cu target, respectively, while increasing input power
increase the sputtering rate from 1.8 A0 s1 to 2.6 A0 s1 for
Cr and 2.5 A0 s1 to 5.6 A0 s1 for Cu. The sputtering yield
of Cr/Cu target to Arþ and Ar (fast) atoms is greater than to
hydrogenic ions. From OES and Langmuir probe study, it is
clear that with addition of hydrogen in pure Argon plasma
leads to decrease of degree of ionization of Ar and corre-
sponding density of Ar like ions. Also, with addition of
FIG. 11. Influence of (a) working pressure (b) input power on deposition
rate for Cr and Cu target.
Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 152.88.96.146 On: Sat, 03 Sep
2016 11:39:27
Phys. Plasmas 20, 103505 (2013)
hydrogen Ar (fast) density decreases linearly (due to Reaction
2, 5 and 9). An exactly opposite trend of the density of Ar like
ions is observed as a function of input power. Due to these
facts, the drop and increase of sputtering rate as a function of
working pressure and input power are seen. It is interesting to
find higher deposition rate for Cu compared to Cr target at
particular deposition condition as sputtering yield(S) of Cu to
Ar(S ¼ 2.3) is higher than Cr(S ¼ 1.3) to Ar.
IV. CONCLUSION
In this paper, the authors have studied the effect of target
on the physical properties of magnetron discharge as a func-
tion of working pressure and input power. For better film dep-
osition plasma parameters like electron temperature, electron
density, degree of dissociation and ionization are key factors.
From the Langmuir probe study, it is found that with addition
of hydrogen in pure Argon plasma, the densities of electrons
and Ar like ions drop gradually. On the other hand, a corre-
sponding increase in Te is observed. From the OES study, a
gradual decline of degree of ionization of Ar as well as degree
of dissociation of H2 as a function of working pressure are
also found. The effect of input power on plasma parameters as
well as on sputtering rate is also examined. From the present
study, it is clear that the addition of H2 in pure Argon plasma
reduces the sputtering rate. Also it is seen that the ISEE coeffi-
cient plays a crucial role in determining plasma parameters
while its effect on sputtering rate of target material is not
observed. The increased value of electron density (Ne), ion
density (Ni), degree of ionization of Ar and degree of dissocia-
tion of H2 for Cr compared to Cu target is explained on the ba-
sis of its higher ISEE coefficient value.
ACKNOWLEDGMENTS
The authors would like to thank Mr. N. Kathar for his
technical support. The authors are gratefully acknowledge to
the Department of Atomic Energy, Govt. of India for the fi-
nancial support to carry out the present work.
1
I. Safi, Surf. Coat. Technol. 127, 203 (2000).
2
P. J. Kelly and R. D. Arnell, Vacuum 56, 159 (2000).
3
A. R. Nyaiesh, Thin Solid Films 86, 267 (1981).
4
A. V. Phelps and Z. L. Petrovic, Plasma Sources Sci. Technol. 8, R21
(1999).
5
J. Musil, L. Bardos, A. Rajsky, J. Vyskocil, B. Dolezal, G. Loncar, K.
Dadourek, and V. Kubicek, Thin Solid Films 136, 229 (1986).
6
J. H. Huang, F. Y. Ouyang, and Ge. P. Yu, Surf. Coat. Technol. 201, 7043
(2007).
7
B. Subramanian and A. K. Jayachandran, Appl. Surf. Sci. 255, 2133
(2008).
8
S. Kondo and K. Nambu, J. Phys. D: Appl. Phys. 32, 1142 (1999).
9
D. Depla, G. Buyle, J. Haemers, and R. D. Gryse, Surf. Coat. Technol.
200, 4329 (2006).
10
R. A. Baragiola, E. V. Alonso, J. Ferron, and A. O. Florio, Surf. Sci. 90,
240 (1979).
11
Y. P. Raizer, Gas Discharge Physics (Springer, New York, 1991).
12
L. M. Kishinevsky, Radiat. Eff. 19, 23 (1973).
13
B. Kakati, S. S. Kaushik, B. K. Saikia, and M. Bandopadhyay, Phys.
Plasma 18, 033705 (2011).
14
W. D. Westwood, S. Maniv, and P. J. Scanlon, J. Appl. Phys. 54, 6841
(1983).
15
G. Buyle, W. D. Bosscher, D. Depla, K. Eufinger, J. Haemers, and R. D.
Gryse, Vacuum 70, 29 (2003).
 103505-9 Saikia, Kakati, and Saikia Phys. Plasmas 20, 103505 (2013)

16 24
A. Bogaerts and R. Gijbels, Spectrochim. Acta, Part B 57, 1071 (2002). M. J. de Graph, R. Servens, R. P. Dahya, M. C. M. van de Sanden, and D.
17
P. F. Knewstubb and A. W. Tickner, J. Chem. Phys. 36, 674 (1962). C. Schram, Phys. Rev. E 48, 2098 (1993).
18 25
M. H. Gordon and C. H. Kruger, Phys. Fluids B 5, 1014 (1993). T. Kimura and H. Kasugai, J. Appl. Phys. 107, 083308 (2010).
19 26
R. F. G. Meulenbroeks, A. J. van Beek, A. J. G. van Helvoort, M. C. M. S. Yugeswaran, K. Suresh, and V. Selvarajan, Plasma Sci. Technol. 12, 35
van de Sanden, and D. C. Schram, Phys. Rev. E 49, 4397 (1994). (2010).
20 27
R. S. Mason, P. D. Miller, and I. P. Mortimer, Phys. Rev. E 55, 7462 (1997). J. B. Shumaker and C. H. Ponenoe, J. Opt. Soc. Am. 57, 8 (1967).
21 28
J. T. Gudmundsson, Plasma Sources Sci. Technol. 8, 58 (1999). W. R. Bennett, Jr., P. J. Kindlmann, and G. N. Mercer, Appl. Opt. Suppl.
22
A. C. Dexter, T. Farrell, and M. I. Lees, J. Phys. D 22, 413 (1989). 2, 34 (1965).
23 29
M. A. Liberman and A. J. Lichtenberg in Principles of Plasma Discharges B. P. Lavrov, A. V. Pipa, and J. R. Opcke, Plasma Sources Sci. Technol.
and Materials Processings, edited by M. A. Liberman and A. J. Lichtenberg 15, 135 (2006).
30
(John Wiley & Sons, Inc., Hoboken, New Jersy, 2005), p. 569. M. Capitelli and M. Dilonardo, Chem. Phys. 24, 417 (1977).

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 152.88.96.146 On: Sat, 03 Sep
2016 11:39:27