FACULTY OF SCIENCE
Diagnostics of High Power Impulse
Magnetron Sputtering Discharge
Habilitation Thesis
JAROSLAV HNILICA
Department of Physical Electronics
Brno, 2023
Bibliographic record
Author: Jaroslav Hnilica
Masaryk University
Faculty of Science
Department of Physical Electronics
Title of Thesis: Diagnostics of High Power Impulse Magnetron
Sputtering Discharge
Academic Year: 2022/2023
Number of Pages: 45
Keywords: magnetron, sputtering, HiPIMS, plasma diagnostics,
optical emission spectroscopy, fast camera,
strip probe, laser induced fluorescence
Abstract
The presented habilitation thesis called Diagnostics of High Power Impulse
Magnetron Sputtering Discharge summarizes and comments on
the results obtained during my work at the Department of Physical
Electronics at the Faculty of Science of the Masaryk University together
with the results attained in collaboration with colleagues from
Academy of Sciences of the Czech Republic in Prague, University of
Liverpool, Paris-Saclay University, National Institute for Optoelectronics
in Bucharest and University of Mons. Eight papers showing a view
on recent observation of the self-organization phenomenon in High
Power Impulse Magnetron Sputtering discharge and spatial and temporal
particle density evolution in the sputtering system were chosen
to be presented in this work.
Acknowledgements
First, I would like to express my gratitude to my wife, Michaela, for
her help and support, without which none of this would be possible.
A special thanks are dedicated to my family and friends.
I would like to thank all my colleagues, present and past, local as
well as from abroad.
Contents
List of commented publications 7
Introduction 9
1 Spokes in High Power Impulse Magnetron Sputtering 11
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.2 The link between the oscillations of voltage and current
induced by spokes . . . . . . . . . . . . . . . . . . . . . 12
1.3 Study of spoke properties in broad range of experimental
conditions . . . . . . . . . . . . . . . . . . . . . . . . 15
1.4 Spoke splitting and merging observed by high-speed
camera and strip probes . . . . . . . . . . . . . . . . . . 18
1.5 In-depth investigation of spoke configuration . . . . . . 20
1.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2 Dynamics of sputtered particles in HiPIMS 24
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.2 Optical diagnostics for characterization of sputtering
process . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.3 Diagnostics of standard and multi-pulse HiPIMS discharge
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.4 Spatial and temporal density evolution determined by
LIF and AAS measurements . . . . . . . . . . . . . . . . 31
2.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Bibliography 37
Author’s publications not discussed in this work 42
6
List of commented publications
I have chosen eight research articles related to the spokes and dynamics
of sputtered particles in HiPIMS discharge as a part of my thesis.
The publications listed below are not ordered by publication date but
according to the topic and as they appear within the following text.
They are denoted as “AP,” which stands for annotated papers. My contribution
to these articles is summarised in the following tables, with
special attention to the experimental work, supervision of students,
manuscript preparation, and research direction.
AP1 KLEIN, Peter; HNILICA, Jaroslav; HUBIČKA, Zdeněk; ČADA, Martin;
ŠLAPANSKÁ, Marta; ZEMÁNEK, Miroslav; VAŠINA, Petr.
Cathode voltage and discharge current oscillations in HiPIMS.
Plasma Sources Sci. Technol. 2017, 26, 055015.
Experimental work Supervision Manuscript Research direction
30% 50% 35% 15%
AP2 HNILICA, Jaroslav; KLEIN, Peter; ŠLAPANSKÁ, Marta; FEKETE,
Matej; VAŠINA, Petr. Effect of magnetic field on spoke behaviour
in HiPIMS plasma. J. Phys. D: Appl. Phys. 2018, 51, 095204.
Experimental work Supervision Manuscript Research direction
25% 100% 50% 90%
AP3 KLEIN, Peter; LOCKWOOD ESTRIN, Francis; HNILICA, Jaroslav;
VAŠINA, Petr; BRADLEY, James W. Simultaneous electrical and
optical study of spoke rotation, merging and splitting in HiPIMS
plasma. Plasma Sources Sci. Technol. 2017, 50, 015209.
Experimental work Supervision Manuscript Research direction
30% 50% 25% 10%
AP4 KLEIN, Peter; HNILICA, Jaroslav; ZEMÁNEK, Miroslav; BRADLEY,
James W; VAŠINA, Petr. The statistics of spoke configurations in
high-power impulse magnetron sputtering discharges. J. Phys.
D: Appl. Phys. 2019, 52, 125201.
7
List of commented publications
Experimental work Supervision Manuscript Research direction
35% 50% 30% 20%
AP5 VAŠINA, Petr; FEKETE, Matej; HNILICA, Jaroslav; KLEIN, Peter;
DOSOUDILOVÁ, Lenka; DVOŘÁK, Pavel; NAVRÁTIL, Zdeněk.
Determination of titanium atom and ion densities in sputter
deposition plasmas by optical emission spectroscopy. Plasma
Sources Sci. Technol. 2015, 24, 065022.
Experimental work Supervision Manuscript Research direction
10% 20% 25% AP6
FEKETE, Matej; HNILICA, Jaroslav; VITELARU, Catalin; MINEA,
Tiberiu; VAŠINA, Petr. Ti atom and Ti ion number density evolution
in standard and multi-pulse HiPIMS. J. Phys. D: Appl. Phys.
2017, 45, 055201.
Experimental work Supervision Manuscript) Research direction
35% 50% 30% 25%
AP7 HNILICA, Jaroslav; KLEIN, Peter; VAŠINA, Petr; SNYDERS, Rony;
BRITUN, Nikolay. Revisiting particle dynamics in HiPIMS discharges.
I. General effects. J. Appl. Phys. 2020, 128, 043303.
Experimental work Supervision Manuscript Research direction
50% - 40% 90%
AP8 HNILICA, Jaroslav; KLEIN, Peter; VAŠINA, Petr; SNYDERS, Rony;
BRITUN, Nikolay. Revisiting particle dynamics in HiPIMS discharges.
II. Plasma pulse effects. J. Appl. Phys. 2020, 128, 043304.
Experimental work Supervision Manuscript Research direction
50% - 40% 90%
8
Introduction
The presented habilitation thesis called Diagnostics of High Power Impulse
Magnetron Sputtering Discharge summarizes and comments on
the results obtained during my work at the Department of Physical
Electronics at the Faculty of Science of Masaryk University. A significant
part of the results was also attained in co-operation with
colleagues from the Czech Republic and abroad, namely from Belgium,
France, Romania, and United Kingdom. These works focus on
non-intrusive spectroscopic discharge characterization techniques. Insitu
non-intrusive discharge diagnostics is more favorable for precise
and explicit discharge characterization as the plasma remains unperturbed
in this case. I have participated in all the work presented in
these studies, focusing on designing the experiments, methodology,
investigation, and sputtered process description.
The presented work is divided into two chapters. The first chapter
deals with plasma self-organization in High Power Impulse Magnetron
Sputtering discharge. In 2012, three independent groups, all using
fast imaging techniques, have simultaneously reported on plasma
self-organization in this type of discharge. They observed that plasma
is not homogeneously distributed over the target under certain experimental
conditions, but it is localized into zones with enhanced
ionization. Additionally, they discovered that these ionization zones
rotate in the E × B direction above the racetrack area of the cathode
(the racetrack being the area on the target with the highest erosion
induced by the sputtering) with velocities of several km/s. Similar
to moving regions of enhanced ionization observed in Hall thrusters
called spokes; the ionization zones are now also called spokes in the
magnetron sputtering community.
The second chapter is dedicated to determining the absolute number
density of sputtered species in the discharge, one of the most
crucial plasma diagnostics tasks. Such knowledge is necessary, especially
the time- and space-resolved dynamics of the sputtered species
in the ground state, to understand the plasma kinetics and chemistry.
The absolute particle density could be used as input data to
model the discharge processes. Moreover, the particle fluxes in the
discharge, which are essential for the thin film growth process, can
9
Introduction
be estimated by knowing the sputtered species density and velocity.
The number densities of sputtered species are determined using the
effective branching fractions method, laser-induced fluorescence, and
atomic absorption spectroscopy.
Both chapters include commentaries of published papers. I commented
on the most important results obtained and the facts that are
not usually widely discussed in published articles, like the motivation
and issues. Note that the papers are not attached to this work because
they are not published in an open-access regime.
10
1 Spokes in High Power Impulse Magnetron Sput-
tering
1.1 Introduction
High power impulse magnetron sputtering (HiPIMS) is a physical
vapor deposition technique utilizing a short high energy pulse (several
10s of µs) followed by a long off-time to deliver the power into
the discharge [1, 2]. It leads to dense plasma formation with a high
fraction of ionized species and consequently high ion flux towards the
substrate. In HiPIMS discharge, it was observed that in some cases,
the plasma is not homogeneously distributed above the target, but
it is concentrated in regions of enhanced ionization, called spokes
[3, 4, 5], see Figure 1.1. The spokes were observed only for a certain
combinations of experimental parameters such as discharge current,
operating pressure, target material, and working gas. The high-speed
camera made it possible to determine the spoke velocity of around
10 km.s−1, the number of spokes, the transition between spoke modes,
and the different characteristic spoke shapes [6]. The same spoke velocity
and spoke rotation frequencies of several hundred kHz were
obtained using electrical probes and optical monitoring [5, 7].
The first chapter of this habilitation thesis focuses on studying the
spoke phenomena in High Power Impulse Magnetron Sputtering. In
the first paper AP1, we described the investigation of the oscillations
on the cathode voltage and the discharge current, which emerged
suddenly and simultaneously during a HiPIMS pulse. Simultaneous
measurement of both oscillations and spokes was conducted to find
mutual correlations. Paper AP2 determined the spoke shape, number,
and velocity for a broad range of experimental conditions in three
magnetic field configurations employing a high-speed camera. Additionally,
the spokes were investigated by optical emission spectroscopy.
Paper AP3 was focused on simultaneous measurements by a highspeed
camera and strip probes embedded in the target. Such a dual
diagnostics system provides for gaining information on spokes temporal
and spatial behavior in HiPIMS plasmas. A correlation between the
optical emission image of an individual spoke and the current deliv-
11
1. Spokes in High Power Impulse Magnetron Sputtering
ered to the target by the spoke has been made for a range of magnetron
operating conditions. Additionally, merging and splitting of spokes
as they rotate has been studied. Paper AP4 investigates quasi-stable
spoke configurations evolution during the HiPIMS pulse by simultaneously
using six embedded strip probes evenly distributed in the target
and by simultaneous observation by a fast camera. This experimental
arrangement tracks the changes in the spoke configuration.
Figure 1.1: Image of the spoke: (left) original grey-scale image from
high-speed camera, (right) image converted to (false) color.
1.2 The link between the oscillations of voltage and current
induced by spokes
Several HiPIMS discharge investigations showed oscillations on both
the cathode voltage and the discharge current at the later stages of the
HiPIMS pulse, where the highest discharge current occurs. Figure 1.2
demonstrates an example of such oscillations on cathode voltage and
discharge current waveforms. In the literature, the oscillations were
detected at various pressures at later stages of the HiPIMS pulse: i)
on the cathode voltage for Ta target [8], ii) discharge current for Ti
target [9], iii) on cathode voltage and discharge current on Nb [10],
LaB6 [11] and Ti [12] targets. Those oscillations manifest only if a
certain target current threshold is reached. This is in line with direct
12
1. Spokes in High Power Impulse Magnetron Sputtering
Figure 1.2: Example of oscillations on (a) cathode voltage and (b)
discharge current waveforms for the pressure of 1 Pa and titanium
target.
observations in the literature, where spokes are only seen above a
certain current level for Ti targets [7] and Cr, Cu, Nb, Mo Ta targets
[13].
Our investigation aimed to verify whether oscillations are connected
to the generator electronics or the spokes. Coincidentally, most
examples of oscillations mentioned above were detected, when a commercial
HiPIMS power supply made by Melec GmbH company was
utilized in the experiments. Because the HiPIMS generator (model
SIPP 2000) made by Melec GmbH company was available in our lab,
we wanted to carry out the experiments with two differently electronically
designed power generators. Therefore, experiments were made
in collaboration with dr. Hubička and dr. Čada from the Institute of
Physics of the Czech Academy of Sciences in Prague, who supplied a
custom-made power generator.
Simultaneous measurement of the cathode voltage and the discharge
current oscillations and spoke formation were conducted to
find mutual correlations and presented in paper AP1. The relation
between the cathode voltage, the discharge current oscillation frequency,
and the spoke frequency was investigated. A sputtering system
equipped with a 20 cm circular titanium target was used. Two different
HiPIMS generators (Melec GmbH and custom made) were employed
to drive the HiPIMS discharge. The Melec generator utilized a coil in
the arc protection circuit, while the custom made generator used a
13
1. Spokes in High Power Impulse Magnetron Sputtering
resistor. A high-speed intensified charge-coupled device (ICCD) camera
PIMAX 3 was applied for discharge imaging. The camera worked
in dual-image feature mode, enabling the capture of two snapshots
during one HiPIMS pulse with only a 3 µs between them. In these and
further experiments, we could determine the spokes shape, number,
and rotation velocity using the PIMAX 3 camera working in this mode.
We wanted to find the conditions where we observe the cathode
voltage and the discharge current oscillations. First, we wanted to find
out if the oscillations are caused only by the electronics of the power
generator. Therefore, we did not connect power generators to the deposition
chamber but specially designed electrical measurement circuits,
and we monitored the V-I characteristics. We did not detect oscillations
in any set conditions covering all the conditions we used in the real
sputtering process. Next, the power generators with the electronics
measurement circuits were connected to the deposition chamber, and
the HiPIMS discharge was ignited. The oscillations were detected only
when an inductance provided by a coil (> 3.2 µH) was a part of the
electrical circuit, and the discharge was sustained. We concluded that
the electronics do not cause the oscillations, but the oscillations are excited
in the plasma and amplified by the electrical circuit in the power
supplies. We found experimental conditions (working pressure and
discharge current) where the oscillations appeared and determined
their frequency.
Additionally, we made a model of the real HiPIMS arrangement, including
the generator circuit and the discharge. The oscillation source
was added to the equivalent electrical circuit representing the discharge.
It was found that the model results agreed well with the measured
data. Thus, the model confirmed that the source of oscillations
is in the HiPIMS plasma.
The fast camera was utilized to capture the spokes for wide range
of experimental conditions. From the images we estimated number of
spokes and rotation velocity of the spoke. We also derived the spoke
rotation frequency from the spoke rotation velocity and racetrack
diameter. In our case, the oscillation frequency was one order of magnitude
higher than the rotation frequency of the spokes. Moreover,
the oscillation frequency and the rotation frequency of spokes showed
different evolution with pressure and the discharge current. Thus, a
direct connection between the rotational motion of spokes and the
14
1. Spokes in High Power Impulse Magnetron Sputtering
oscillations on the cathode voltage and the discharge current was not
found. Nevertheless, we observed that oscillations always accompanied
the spokes because we never detected the oscillations without
spokes.
1.3 Study of spoke properties in broad range of experimental
conditions
The results in the literature on the spoke shape [13, 14, 15, 16, 17], number
[3, 5, 7], and rotation velocity [5] suggest that the spoke behavior
depends on the experimental configuration. Therefore, comparing
the results obtained from two different laboratory devices is difficult,
even if the conditions, such as pressure and discharge current density,
are similar. We realized this issue while writing the previous article
AP1. Therefore, paper AP2 aimed to look at the effect of experimental
conditions such as working pressure and discharge current density on
spoke appearance and behavior, emphasizing the role of the magnetic
field strength.
We used a 7.6 cm circular titanium target and intended to cover
broad range of experimental conditions. The working pressure was
investigated in the range of 0.1 – 5.0 Pa. It covers the whole range of
pressures used for sputter depositions. The applied discharge current
density was up to 10 A/cm2, which was the limit of the power supply.
In this work, three different magnetic field strengths were applied.
Such investigation was inspired by previous works [18, 19], where the
weakening of the magnetic field strength led to a significant increase
in the deposition rate, which was explained by the decrease of the
backward flux of ions to the target [19]. The change of the target
thickness modified the magnetic field strength at the surface, see
Figure 1.3. The PIMAX 3 high-speed camera with dual image feature
enabled us to determine the spokes shape, number, and velocity.
Until now, most reports described only the triangular and diffusive
spoke [16]. We captured thousands of plasma emission images
with a fast camera. All images were processed manually, and several
spoke shapes were recognized. Therefore, developing a methodology
to distinguish different spoke shapes was necessary. The criteria
15
1. Spokes in High Power Impulse Magnetron Sputtering
Figure 1.3: Magnetic field configuration map with depicted positions
of the targets with different thicknesses.
for determining spoke appearance were developed semi-empirically
using dual images focusing on spoke shapes and spoke motion:
1. Non-recognizable spoke – plasma emission is homogeneously
distributed without plasma self-organisation.
2. Stochastic spokes – the plasma is self-organized into the rotating
spokes without noticeable periodicity, and well-defined spoke
borders.
3. Triangular spoke – a narrow trace progressively expanding in
width and intensity until a sharp drop of emission signal terminates
it. Spokes are uniformly distributed over the racetrack.
4. Diffusive spoke – spokes are elongated and dispersed, with a
narrow trace start, the broad middle part of the spoke, and a
narrow trailing edge observed. Spokes are uniformly distributed
over the racetrack.
5. Round spoke – spoke exhibits round shape with neither the
sharp end in plasma emission as triangular spoke nor elongated
16
1. Spokes in High Power Impulse Magnetron Sputtering
dispersed appearance as diffusive spokes. The spokes are uniformly
distributed over the racetrack.
The typical image of different spoke shapes is shown in Figure 1.4.
For each magnetic field, a map was created relating the spoke shape
occurrence with the discharge conditions. In literature, we can find
contradicting results, i.e., an increasing number of spokes [3] as well
as the decreasing number of spokes [7] with increasing discharge
current. Thus, we decided to investigate this issue and find out what
was correct. The answer is that both previously reported results are
right. We revealed that both the increase and the decrease of the
number of spokes with increasing the discharge current could be
reached by variation of experimental parameters. Additionally, we
reported that the spoke rotation velocity decreased by increasing the
pressure and the magnetic field strength. Thus, the number of spokes
and spokes velocity were found to strongly depend on the discharge
conditions.
Figure 1.4: Images of characteristic plasma emission.
Our experimental results contradicted the early reported hypothesis
[16] that titanium as the target material with the lower second
ionization potential should exhibit only a diffusive spoke shape. Both
triangular and diffusive spoke shapes were observed for the titanium
target in our case. The optical emission spectroscopy revealed that
triangular spokes correlated with the strong argon ion emission. The
round spokes presence was connected to the decreasing emission of
17
1. Spokes in High Power Impulse Magnetron Sputtering
a titanium ion with increasing discharge current due to the creation
of multiply charged ions. We proposed a model which describes the
spoke shapes based on these findings. In our model, a triangular spoke
shape is formed due to secondary electron production induced by
argon ions from the spoke sides. The magnetic field then traps these
electrons, forming a distinct trailing edge. Round spokes are created
by the secondary electrons generated within the spoke by the impact
of multiple ionized metal ions. Mapping the spoke properties led to
their systematic description, which had been missing until then. Additionally,
we used newly gained information about spokes to improve
the spoke model used so far.
1.4 Spoke splitting and merging observed by high-speed
camera and strip probes
Group of prof. Bradley from Liverpool University detected periodic
oscillations in the local target current density by electrical isolated
(strip) probes embedded in the target and attributed these oscillations
to the presence of rotating spokes [20]. However, visual confirmation
of spokes presence was not possible in their arrangement. Since the
high-speed camera proved to be a powerful instrument for spoke investigation,
we carried out joint experiments, where we wanted to
confirm that observed strip probe current modulations are indeed
due to spokes. Therefore, simultaneous electrical and optical measurements
in HiPIMS plasma were performed in a joint investigation with
prof. Bradley and his Ph.D. student in our lab.
In this study AP3, we employed a 7.6 cm circular niobium target.
The strip probes embedded in the target enabled us to observe a distinct
modulation in the local target current density as the spokes pass
over a particular target region. Three strip probes, made of the target
material, were placed in machined slots at three angular positions
of 25◦ from each other around the target, see Figure 1.5. They are
always at the same potential as the target, which makes them indistinguishable
for the plasma from the rest of the target. The strip probes
extended beyond the target edge to allow electrical connections to the
power supply. A thin layer of polyimide tape was wrapped around
three sides of the stripe probes to isolate them electrically from the
18
1. Spokes in High Power Impulse Magnetron Sputtering
Figure 1.5: Experimental arrangement of the strip probes from the top
view.
rest of the target. This is tricky because the tape is very thin, and the
space between the probe and the target is tight. Additionally, the tape
should not rise up, and the strip probes should be at the same level as
the target to prevent arcing on the sharp edges. Note that discharge
current reached up to 240 A, i.e., 5.3 A/cm2. Thus, correctly placing
the embedded probes was critical to the entire experiment. Simultaneously,
we detected the plasma emission from the top by high-speed
camera PIMAX 3 .
We revealed that high-speed camera imaging and embedded strip
probes in the target show identical profiles of passing spoke structures
with a sharp trailing edge, see Figure 1.6. Such a diagnostic system
allows us to get information about the discharge above the whole
target surface by camera imaging twice per HiPIMS pulse and simultaneously
get data from three localized places above the target using
embedded strip probes during the same HiPIMS pulse. We utilized
this experimental setup consisting of the camera and strip probes
to investigate time-resolved spoke dynamics. We revealed that the
number of spoke changes during the single pulse. For the first time,
we captured both the merging and splitting of a set of spokes as they
rotate. After merging or splitting events, the new spoke configuration
was not always stable in time. During these measurements, a large
19
1. Spokes in High Power Impulse Magnetron Sputtering
spoke could be divided into two smaller spokes only to merge a short
time later. These observations confirmed that the spokes are a dynamic
phenomenon.
Figure 1.6: 2D plasma emission image (a) of the whole target taken by
high-speed camera at 191 µs with positions of the strip probes (red,
green and blue) and (b) the corresponding strip probe waveforms for
working pressure of 0.14 Pa. One spoke rotation period is illustrated
by the grey background.
Furthermore, we developed a simple phenomenological model,
which relates the stable configuration of spokes to their dimensions, rotation
velocity, and the background Ar gas atoms velocity. The model
estimates the number of spokes that can be sustained for particular
plasma conditions. The model still holds because no contradicting
report occurs in the literature. An in-depth study of the spoke configuration
was performed later and discussed in the text hereafter.
1.5 In-depth investigation of spoke configuration
The results in paper AP3 revealed that as the spokes rotate above the
target, their properties can change over time, with frequently occurring
splitting or merging. Applying our knowledge about spoke shape and
configuration gained from earlier experiments AP2, we further extend
all these investigations as we intended to study the changes in spoke
configuration more deeply AP4.
20
1. Spokes in High Power Impulse Magnetron Sputtering
In this study AP4, the cooperation with prof. Bradley continued.
We planned to utilize a niobium target with six embedded strip probes
evenly distributed over the target, as shown in Figure 1.7(a), together
with the high-speed camera. Such probe arrangement leads to tracking
the changes in the spoke configuration around the racetrack at any
moment in time. However, the standard anode prevents the installation
of the strip probes. We made a few measurements without any anode,
but we were not satisfied with the stability and repeatability of the
experiments. Therefore, we designed a custom-made anode elevated
a few millimeters above the target surface, which could be utilized
with the strip probes, see Figure 1.7(b). Based on our previous paper
AP2, we chose to study (a) the low-pressure case (0.2 Pa), where
the triangular spoke should be present, and the high-pressure case
(4 Pa), where the round spoke should occur. A 2.4 Ω resistor integrated
into series with a cathode worked as negative feedback limiting the
runaway of the discharge current. Thus it increased the stability of the
discharge current.
Figure 1.7: Experimental arrangement from the top view (a) and the
side view (b) on the slotted target outfitted with six strip probes.
Since we had signals from 6 localized places above the target, it
was necessary to get information about the spoke configuration from
the rest of the target. At first, all six strip probe current signals were
processed to recreate the spoke configuration at any point of time
during the HiPIMS pulse. After that, the spoke configuration rotational
velocity was determined during the pulse at the positions of the probes
using a processing script. Afterward, the rotation velocity temporal
21
1. Spokes in High Power Impulse Magnetron Sputtering
evolution was used to transform the data to a frame rotating with
the spokes. Virtual data points between the probes were found from
interpolation. An example of a reconstructed spoke configuration is
illustrated in Figure 1.8, where the color scale from blue to red is
adjusted for the current minima and maxima, respectively. The figure
shows us from 65 to 135 µs seven spokes. At 135 µs, we can detect the
merging of two spokes and simultaneously the splitting of one spoke.
Figure 1.8: An example of the reconstructed spoke configuration during
the pulse from 50 to 150 µs at the pressure of 4 Pa.
We recorded around 150 pulses under the same experimental conditions
to obtain sufficient statistical data. We identified several possible
scenarios of spoke behavior during the HiPIMS pulse using images
with reconstructed spoke configuration. The merging and splitting
events occurred in various combinations on different target parts,
either successively or seemingly simultaneously. The merging and
splitting of spokes was discovered to influence the neighboring spokes.
For the low-pressure case (0.2 Pa), the most probable configuration
was four spokes, with an equal probability for spokes to merge and
split. In the high-pressure case (4 Pa), the spoke configuration exhibited
a higher number of spokes, i.e., 7 or 8 spokes. An increasing
number of spokes with increasing pressure corroborates well with our
22
1. Spokes in High Power Impulse Magnetron Sputtering
previously reported results using titanium target AP2. Furthermore,
we observed a significantly higher occurrence of spoke splitting versus
spoke merging during a single pulse.
We also created a model that includes the stable spoke configuration,
spoke splitting, and spoke merging. Anders phenomenological
model [14] served as an inspiration for us. We assumed: i) electrons
rotating in E × B ionised sputtered metal, ii) metal ions are backattracted
to the target, and iii) re-sputtering. We described the spatial
evolution of electrons, Nb atoms, and Nb ion concentration in the stable
spoke in the model. Then we showed how the small perturbation
in Nb atom concentration could evolve in time, and a new spoke is
formed. Thus, we described how the spoke can split. Similarly, the
model can explain the spoke merging in the reverse case.
1.6 Conclusion
This section was devoted to the investigation of the spoke phenomenon
using optical emission spectroscopy, high-speed camera, and strip
probes. An extensive study of spoke images for various pressures,
discharge currents, and magnetic fields led to the development of
a methodology that allowed us to categorize spoke based on its appearance.
The results from strip probes showed us how dynamic phenomenon
spokes are and that a statistical approach is needed to study
them. In the future measurements, the experimental data describing
the plasma parameters inside the spokes are required to better understand
the spoke phenomenon. At the same time, due to the high
power load, all diagnostics must be non-invasive, making the spoke
investigation still a big challenge.
23
2 Dynamics of sputtered particles in High Power
Impulse Magnetron Sputtering
2.1 Introduction
A deep understanding of the physical processes driving the HiPIMS
discharge is essential for optimizing thin-film growth and developing
even more efficient sputtering processes with an enhanced level of
discharge control. Only a few techniques suitable to follow the temporal
and spatial evolution of sputtered species densities have been
developed, such as atomic absorption spectroscopy or laser-induced
fluorescence [21, 22].
We want to use a simpler method to determine the density of
sputtered species as well as the method that does not disturb the
plasma. However, the sputtered species in the ground or metastable
states do not emit any direct optical signal, and it is difficult to obtain
information about the number densities of the species in these levels
from optical emission spectroscopy. Thus, we developed an indirect
method based on self-absorption [23, 24] to determine the number
densities of sputtered species in the discharge.
This chapter concentrates on the study of HiPIMS processes utilizing
the effective branching fraction (EBF) method, laser-induced fluorescence
(LIF) imaging, and atomic absorption spectroscopy (AAS).
The methods mentioned above have been used to determine the ground
state number density of Ti atoms and ions. In paper AP5, we applied
the effective branching fractions method to determine the ground state
densities of the sputtered titanium atoms and ions and the evolution
of titanium atoms and ion densities in three cases ranging from the
direct current (dc) driven sputter process to HiPIMS. The following
paper AP6 deals with the density evolution of sputtered species determined
by EBF in single and multi-pulse HiPIMS. Moreover, the results
obtained by the EBF method were compared to results determined using
the tunable diode-laser absorption spectroscopy (TD-LAS) in the
dc case. Papers AP7 and AP8 were published simultaneously. These
extended and detailed experimental studies of HiPIMS processes were
performed by time-resolved imaging of the ground state sputtered
24
2. Dynamics of sputtered particles in HiPIMS
particles. Paper AP7 focuses on the role of main discharge parameters,
such as pulse repetition rate and pulse energy. Paper AP8 deals with
the effects of the plasma-on phase for different plasma pulse duration,
pulse energy, gas pressure, and molecular oxygen admixture.
2.2 Optical diagnostics for characterization of sputtering
process
There are plenty of non-radiating species in the ground states in sputtering
discharges. It is challenging to acquire the number densities of
the species in these states from optical emission spectroscopy. But how
do we measure them non-destructively and easily? A method based on
self-absorption has been invented to estimate the number densities of
argon [25, 26] and neon [26, 27] metastable states in low-temperature
plasma. We adopted this method using effective branching fractions
and determined the density of non-radiating species from the intensities
of self-absorbed spectral lines.
In paper AP5, we introduced the effective branching fractions
(EBF) method to determine the sputtered titanium atoms and ions
ground state densities. The procedure is based on fitting the theoretically
calculated branching fractions to experimentally measured ratios
of the relative intensities of selected lines of the measured species. At
first, it was necessary to choose the suitable candidates for the titanium
atom and ion lines from the NIST database [28], fulfilling the
condition of one transition going to the ground state. Then we have to
find these lines in an overview spectrum. The lines should be easily
measured and not overlap with surrounding spectral features in the
measured spectra. Additionally, they have to exhibit a high transition
probability (Einstein coefficient). An example of the branching fraction
fit used to determine Ti atom density is presented in Figure 2.1.
In this case, the best fit of theoretically calculated branching fractions
to the experimentally measured ratios of the relative intensities is
achieved when the Ti atom density is 1×1017 m−3. Such a result is
comparable with the density reported in the literature [29].
The manuscript AP5 also presented the lists of carefully selected Ti
atom and Ti ion lines applied for density determination. However, it
is not necessary to use all the lines summarized in Table I and Table II
25
2. Dynamics of sputtered particles in HiPIMS
in the paper AP5. For example, we took 13 lines of Ti atoms and 19
lines of Ti ions and utilized them for the several case studies, where
we applied EBF method to estimate:
1. The titanium atom density dependence on the power and working
pressure in a dc magnetron sputtering discharge.
2. Titanium atom and ion densities for the transition from dc to
HiPIMS regime.
3. Temporal evolution of titanium atom and ion densities during
short and long HiPIMS pulse.
Figure 2.1: The effective branching fraction fit used to determine Ti
atom density. Calculation was done for z = 5 cm, Texc = 3000 K,
Tgas = 300 K.
For the measurements of the first two cases, we utilized a chargecoupled
device (CCD) detector, while for the temporal evolution of
26
2. Dynamics of sputtered particles in HiPIMS
sputtered species in the HiPIMS pulse, we applied an ICCD detector.
The relative uncertainty of density determination was about 5%
for measurements made by the CCD detector. In contrast, for timeresolved
measurements made by the ICCD detector, it was around
15% in our experimental setup. Note that these uncertainties are determined
statistically by the fitting procedure from the residual sums
of squares.
We measured the titanium atom and ion lines and determined
the densities for all three case studies. It was difficult to compare the
densities in some cases since different experimental setups, magnetic
fields, and in the case of HiPIMS, different peak discharge currents or
pulse lengths were used. Nevertheless, determined Ti atom and ion
densities and the evolution of Ti atom density with the applied power
and pressure were comparable to those in the literature, proving that
the EBF method gives reliable data in sputtering discharges. The EBF
method does not require any reference source, i.e., a lamp or a diode
laser. The Ti atom and ion densities are determined directly from optical
emission spectroscopy signals. The OES measurement is simple
and enables us to implement the EBF method into any sputtering experiments.
Only a vacuum feed-through or a window for optical fiber,
together with a spectrometer with sufficient resolution, is required
for the measurement. Thus, the EBF method enables investigation of
the time-resolved dynamics of sputtered species simply, reliably, and
without plasma disturbing.
2.3 Diagnostics of standard and multi-pulse HiPIMS
discharge
The main disadvantage of HiPIMS is a low deposition rate compared
to the direct current magnetron sputtering (dc-MS) for an equivalent
average applied power [30] and, consequently, a longer process
for the same thickness when this technique is used on an industrial
scale. Therefore, a deposition rate increase would make the deposition
process more economical and efficient, while the deposited thin film
properties would be maintained or even enhanced compared to dc-MS.
Recently, several single-pulse HiPIMS (s-HiPIMS) were put close to
each other, creating a package of pulses [31]. Thus the HiPIMS was op-
27
2. Dynamics of sputtered particles in HiPIMS
erated in the so-called chopped [31], burst [11], or multi-pulse regime
(m-HiPIMS) [32]. It was demonstrated that the deposition rate in mHiPIMS
could be significantly higher for the same average power than
in s-HiPIMS [33]. The tailoring of the multi-pulse sequences allows
for further tuning of the deposition process. Using the m-HiPIMS, the
first results showing changes in coating structure and characteristics
have already been published [33, 34].
The initial results were reported mainly on the thin film deposition
by m-HiPIMS with limited plasma diagnostics and an understanding
of the sputtering process. Therefore, we decided to investigate
the multi-pulse HiPIMS by the EBF technique in this paper AP6. Ti
atom and ion densities determined in m-HiPIMS were compared to
the densities in a single-pulse HiPIMS. Moreover, we compared the
results acquired by the EBF method with those obtained from the
well-established TD-LAS method. TD-LAS is a sophisticated and expensive
method that requires an experienced researcher. Therefore,
we conducted the experiments in the lab of prof. Minea at the University
Paris-Saclay in France, in cooperation with dr. Vitelaru from the
National Institute for Optoelectronics in Romania.
A cylindrical vacuum chamber equipped with a 10 cm circular
titanium target was used in these experiments. Simultaneous measurements
of TD-LAS and EBF methods were conducted in the same
investigated volume. The number densities of Ti atom were measured
as a function of the discharge current and the working pressure. It
was demonstrated that Ti atom number densities determined by the
EBF method and TD-LAS agree throughout the investigated conditions.
The agreement between the EBF method and TD-LAS technique
results can be taken as the validation of the EBF method by another
independent measurement. The EBF method proved to be a reliable
technique for the Ti atom number density determination. A similar
comparison was not performed for Ti ions due to the low signal of
Ti ion lines. However, because the EBF method is independent of the
charge and the calculation procedure principle is the same for Ti atoms
and Ti ions, the method can also be used in Ti ion number density
measurements.
The multi-pulse HiPIMS investigation was performed using the
titanium target with a diameter of 20 cm. The study was focused on the
simplest case, where the pulse sequence was composed of two pulses.
28
2. Dynamics of sputtered particles in HiPIMS
Figure 2.2: The discharge current, the number density of Ti atoms and
Ti ions in m-HiPIMS discharge. The delay between the first and the
second pulse was set to 200 µs. The operation pressure was set to 5 Pa,
the pulse duration of both pulses to 200 µs and the period to 100 ms.
The pulse duration was set to 200 µs with discharge current peaking
around 350 A (1.1 A/cm2), because we could easily detect titanium
atoms and ions for such parameters as reported in our previous paper
AP5. The delay time between pulses in the pulse sequence was set to
200, 400, and 1500 µs. We expected that the first pulse would influence
the second pulse in the pulse sequence because remaining charged
carriers will make discharge ignition in the second pulse easier. Additionally,
the sputtered species from the first could also be available
for ionization during the second pulse. The total number of pulses
per second was kept constant as we planned to compare the results
obtained from s-HiPIMS and m-HiPIMS.
The Ti atom and ion density evolution showed that the residual titanium
atoms and ions from the first pulse are crucial for the initiation
and development of the second pulse, see Figure 2.2. The residual titanium
atoms produced by the first pulse are already thermalized at the
beginning of the second pulse and can be ionized during the second
pulse. The discharge current rises faster in the second pulse but does
not reach the same maximum as the first pulse, see Figure 2.2. The
time evolution of the titanium atom density showed different behavior.
29
2. Dynamics of sputtered particles in HiPIMS
The initial increase is followed by a decrease in the first pulse, while a
relatively constant progression during the second pulse is observed.
The Ti ion number density steadily increased during both the first and
second pulse, while the increase was steeper in the first pulse.
For the delay of 400 µs, the evolution of Ti atom and ion was similar
as for the 200 µs delay. Only the reservoir of thermalized atoms gets
depleted as the delay increases. Surprisingly, we still detected the
sputtered Ti atom for the delay of 1500 µs, and the evolution of the Ti
atom during the second pulse was similar to shorter delays. For this
longest investigated delay, Ti ion evolution was the same as in the case
of the first pulse.
From the deposition point of view, we found out that the ionization
fraction of sputtered species is still very high in the m-HiPIMS case
reaching always more than 75%. The total ion flux to the substrate
increased by more than 40% in m-HiPIMS with 200 µs delay compared
to s-HiPIMS. However, we must keep in mind that we measured the
overall ion flux of Ti ions and Ar ions. We can speculate that the
delay of 200 µs is not long enough for Ar gas to replenish, and this
increase could be attributed mostly to the Ti ions. This could be quickly
resolved by additional measurements using the biasable quartz crystal
monitor, which can measure total ion and atom flux or total atom flux
of titanium and thus determine the role of titanium ions. Additionally,
at least 20% increase in deposition rate was detected in m-HiPIMS
compared to s-HiPIMS, which is in agreement with results reported
in literature [33].
The above mentioned results showed that the delay between pulses
in the pulse sequence plays a crucial role in ion flux to substrate enhancement.
Additionally, using the m-HiPIMS instead of a singlepulse
HiPIMS led to an increase in the deposition rate and thus increased
HiPIMS technology efficiency. Since we studied the simplest
multi-pulse HiPIMS case, there is a potential for further improvements.
On the other hand, more parameters make the whole sputtering process
more complicated and open more questions, such as what pulse
duration, the delay between the pulses, or the number of pulses should
be the best or optimal for a particular deposition process.
30
2. Dynamics of sputtered particles in HiPIMS
2.4 Spatial and temporal density evolution determined by LIF
and AAS measurements
During our previous works AP5, AP6, we realized that comparing experimental
results such as sputtered particle densities in the literature
is difficult and sometimes not even possible. The reason is that different
laboratories use different experimental setups, magnetic field configurations,
pulse lengths, repetition rates, working pressures, etc. Until
2020, an overview article, which covers the time and spatially-resolved
density evolution of sputtered species in a broad range of experimental
conditions, did not exist. Therefore, we intended to systematically investigate
the time-resolved ground state sputtered particles in HiPIMS
discharge in a different position from the target surface. For this complex
research, we established a collaboration with dr. Britun and prof.
Snyders from the University of Mons in Belgium. They had experience
with two-dimensional time-resolved density mapping of the ground
state titanium neutrals and ions by laser-induced fluorescence (LIF)
[21, 22]. In our joint investigation AP7, AP8, we focused on the role
of main discharge parameters such as pulse duration, repetition rate,
plasma pulse energy, gas pressure, and molecular oxygen admixture
effects. The behavior of neutral and singly ionized atoms were studied
during the plasma-on and plasma-off time phases.
The experiments were carried out in the lab of prof. Snyders at
the University of Mons in Belgium. Two balanced magnetron sources
equipped with circular planar titanium targets, one of 7.6 cm in diameter
and the other of 10 cm in diameter, have been used. For diagnostics,
we utilized LIF imaging for two-dimensional time-resolved density
mapping of the ground state titanium neutrals and ions together with
AAS.
It is desirable to gather the information from the proximity of
the target surface. However, the standard anode prevents observing
the ionization region, an essential part of the discharge, where the
sputtered species are ionized. Thus, we adopted our custom-made
anode utilized in the experiments with strip probes AP4. Moreover,
we made the anode thinner (about 1 mm) to keep the shielded area
as small as possible, see Figure 2.3.
31
2. Dynamics of sputtered particles in HiPIMS
Figure 2.3: Magnetron head configuration with custom-made anode.
For part of the investigation in paper AP7, we utilized this newly
designed anode geometry, which allowed the laser beam to touch
of the cathode surface. It enables particle visualization close to the
target surface as shown in Figure 2.4, where the typical example of the
ground state density distribution for sputtered Ti neutrals and ions
above the magnetron target in the dc and HiPIMS case is presented.
The ground state density is calibrated by the AAS method. We chose
a logarithmic density representation for Ti neutral and ion density.
In this representation, a single color step corresponds to a threefold
density change. It enables us to visualize the particle behavior up to
10 ms after the end of the HiPIMS pulse, where densities are three
orders of magnitude lower than in plasma-on time. The left column
in Figure 2.4 shows a homogeneous distribution of Ti atoms (upper
image) and a very low density of the Ti ions. On the right upper
image in Figure 2.4, we can observe that ground-state neutral atoms
being ejected from the target during the plasma-on time with maximal
density at 1 – 2 mm above the target surface. As the sputtered titanium
atoms enter the ionization zone, i.e., moving away from the target,
they are ionized, and we can observe a significant reduction in the
titanium atom density. The depletion zone coincides with higher Ti
ions density, as demonstrated in the lower right image. The ionization
zone is identified in these two papers AP7, AP8 as the area above the
target where electrons are confined, creating titanium ions.
32
2. Dynamics of sputtered particles in HiPIMS
Figure 2.4: Ground state density distributions for Ti neutrals and Ti
ions measured above the magnetron target for dc and HiPIMS case.
Data were averaged over 30 laser pulses. The pressure was set to 2.7 Pa.
The log-color scale is used.
Our motivation for investigating the repetition rate effect comes
from our previous paper AP6, where we discovered that the second
pulse in the pulse sequence could be influenced by the first pulse
even when 1500 µs delay between them is applied. As AAS and LIF
measurements enable us to study plasma-on, and plasma-off time,
i.e., they were ideal for such research. We observed that increasing
the repetition rate from 50 to 400 Hz leads to an increased residual
density of the sputtered species, which agrees with our previously
reported results AP6. At the same time, the space-averaged ionization
fraction remains the same regardless of the plasma repetition rate. The
ionization gets uniform in the studied region after about 1500 µs from
33
2. Dynamics of sputtered particles in HiPIMS
the pulse beginning, no matter the pulse repetition rate, which implies
a fast particle thermalization.
We also investigated the particle propagation dynamics using the
vector diagram approach, representing the density gradients for both
sputtered neutrals and ions along with their temporal changes. This
approach confirmed the idea of in-flight ionization of the sputtered
neutrals. The same order of magnitude values was found for neutrals
and ions at the beginning of plasma pulse, nevertheless having an
opposite sign.
In paper AP7, we also compared the titanium atom and ion number
density gained by the AAS method with the results obtained from
the EBF method. An agreement between results obtained by these
two techniques (AAS and EBF) has been found. At this point, the
EBF method was cross-checked by absorption spectroscopy in two
different labs showing that the EBF method is a reliable technique
for determining titanium atom and ion number density in magnetron
sputtering discharge.
The missing information in this work is the density evolution of
argon ions, as they play a crucial role in the sputtering of the target.
Based on the literature, we tested different excitation wavelengths by
the LIF. Despite the great effort, we did not successfully detect any
signal from argon ions. Therefore, future work should focus on this
subject.
Furthermore, we examined the influence of plasma pulse duration,
pulse energy, gas pressure, and molecular oxygen admixture by LIF
technique, focusing on the plasma-on time AP8. This paper was published
simultaneously with the paper AP7, and we considered it its
extension. In this study AP8, only the anode covering the ionization
region, i.e., the area close to the target surface, was used. For each
experiment, we created time-resolved 2D number density maps of
titanium atoms and titanium ions. In addition to the density maps,
the ionization fraction (IF) maps have been built for all studied conditions.
The IF has been calculated based on the calibrated LIF data by
the AAS technique. We also further processed the LIF images to get
additional information. After applying the Abel transformation, the
space-averaged density of the titanium atoms and titanium ions were
calculated from the imaged region.
34
2. Dynamics of sputtered particles in HiPIMS
Since we investigated a broad range of experimental conditions a
short summary of the most important result follows:
• LIF mapping shows the ground state density evolution, both for
neutrals and ions.
• The maximum ionization corresponds to the shortest applied
pulse for the same applied pulse energy.
• The ion production may increase by a factor of 2.5, while the
neutral density may drop 1.5 times when the pulse energy is
doubled.
• The area above the target surface corresponding to high IF values
shrinks with increasing argon pressure.
• An order of magnitude lower density of neutrals and ions is
observed as a result of only a fourfold pressure decrease.
• In the poisoned regime, a decrease in titanium neutral and ion
ground state densities was seen throughout the investigated
region of interest. Titanium ion density is decreased 30 times
with a 10% oxygen admixture.
• The ionization fraction itself is found to be higher in the poisoned
regime, while the spatial distributions of the sputtered atoms in
both reactive and non-reactive cases are very similar.
These two joint papers AP7, AP8 enable us to compare the particle
dynamics in very different experimental conditions, which was the
goal of this research. Additionally, we showed that using a custommade
anode allowed us to observe and investigate the area directly
above the target (ionization region), where the atoms are sputtered
from the target and consecutively ionized. Until now, the models
only theoretically investigated this ionization region. Thus, the experimental
results presented in paper AP7 could be further utilized
to verify and improve the ionization region model [35] dealing with
time-dependent volume-averaged plasma chemical description of the
ionization region in close vicinity of the target racetrack.
35
2. Dynamics of sputtered particles in HiPIMS
2.5 Conclusion
This section presented a simple, non-invasive optical method for studying
the magnetron sputtering process. This method, called the effective
branching fraction method, enabled us to determine the density of
titanium atoms and titanium ions. We applied this method to direct
current magnetron sputtering, single pulse and multi-pulse HiPIMS
plasma. The results of this method coincided both quantitatively and
qualitatively with the results from the literature and the results from
atomic absorption spectroscopy. Afterward, we studied the spatial
and temporal density of sputtered species from the target to the substrate
using the EBF method as well as LIF and AAS techniques. These
investigations brought essential information that contributed to the
understanding of the HiPIMS discharge. Moreover, the results from
multi-pulse HiPIMS showed that such an approach could be suitable
for sufficient ionization of the sputtered particles at an adequately
high deposition rate.
36
Bibliography
1. HELMERSSON, Ulf; LATTEMANN, Martina; BOHLMARK, Johan;
EHIASARIAN, Arutiun P; GUDMUNDSSON, Jon Tomas.
Ionized physical vapor deposition (IPVD): A review of technology
and applications. Thin Solid Films. 2006, vol. 513, no. 1, pp. 1–
24.
2. GUDMUNDSSON, Jon Tomas. The high power impulse magnetron
sputtering discharge as an ionized physical vapor deposition
tool. Vacuum. 2010, vol. 84, no. 12, pp. 1360–1364.
3. KOZYREV, Andrey Vladimirovich; SOCHUGOV, N. S.; OSKOMOV,
Konstantin Vladimirovich; ZAKHAROV, Aleksandr Nikolaevich;
ODIVANOVA, A. N. Optical Studies of Plasma Inhomogeneities
in a High-Current Pulsed Magnetron Discharge. Plasma
Physics Reports. 2011, vol. 37, pp. 667–673.
4. ANDERS, André; NI, Pavel; RAUCH, Albert. Drifting localization
of ionization runaway: Unraveling the nature of anomalous
transport in high power impulse magnetron sputtering. Journal
of Applied Physics. 2012, vol. 111, no. 5, p. 053304.
5. EHIASARIAN, Arutiun P.; HECIMOVIC, Ante; DE LOS ARCOS,
Teresa; NEW, R.; SCHULZ-VON DER GATHEN, Volker; BÖKE,
Marc; WINTER, Jörg. High power impulse magnetron sputtering
discharges: Instabilities and plasma self-organization. Applied
Physics Letters. 2012, vol. 100, no. 114101, p. 114101.
6. HECIMOVIC, Ante; VON KEUDELL, Achim. Spokes in high
power impulse magnetron sputtering plasmas. Journal of Physics
D: Applied Physics. 2018, vol. 51, no. 45, p. 453001.
7. WINTER, Jörg; HECIMOVIC, Ante; DE LOS ARCOS, Teresa;
BÖKE, Marc; SCHULZ-VON DER GATHEN, Volker. Instabilities
in high-power impulse magnetron plasmas: from stochasticity to
periodicity. Journal of Physics D: Applied Physics. 2013, vol. 46, no.
8, p. 084007.
37
BIBLIOGRAPHY
8. BOHLMARK, Johan; HELMERSSON, Ulf; VANZEELAND, Michael;
AXNÄS, Ingvar; ALAMI, Jones; BRENNING, Nils. Measurement
of the magnetic field change in a pulsed high current
magnetron discharge. Plasma Sources Science and Technology. 2004,
vol. 13, no. 4, p. 654.
9. ANDERSSON, Joakim; EHIASARIAN, Arutiun P; ANDERS, André.
Observation of Ti 4+ ions in a high power impulse magnetron
sputtering plasma. Applied Physics Letters. 2008, vol. 93,
no. 7, p. 071504.
10. ANDERS, André. Discharge physics of high power impulse magnetron
sputtering. Surface and Coatings Technology. 2011, vol. 205,
S1–S9.
11. FRANZ, Robert; CLAVERO, César; BOLAT, Rustem; MENDELSBERG,
Rueben; ANDERS, André. Observation of multiple charge
states and high ion energies in high-power impulse magnetron
sputtering (HiPIMS) and burst HiPIMS using a LaB6 target.
Plasma Sources Science and Technology. 2014, vol. 23, no. 3, p. 035001.
12. MASZL, Christian; BREILMANN, Wolfgang; BENEDIKT, Jan;
VON KEUDELL, Achim. Origin of the energetic ions at the substrate
generated during high power pulsed magnetron sputtering
of titanium. Journal of Physics D: Applied Physics. 2014, vol. 47, no.
22, p. 224002.
13. DE LOS ARCOS, Teresa; LAYES, V; GONZALVO, Y Aranda;
SCHULZ-VON DER GATHEN, Volker; HECIMOVIC, Ante; WINTER,
Jörg. Current–voltage characteristics and fast imaging of
HPPMS plasmas: transition from self-organized to homogeneous
plasma regimes. Journal of Physics D: Applied Physics. 2013, vol. 46,
no. 33, p. 335201.
14. ANDERS, André. Self-organization and self-limitation in high
power impulse magnetron sputtering. Applied Physics Letters.
2012, vol. 100, no. 224104, p. 224104.
15. ANDERS, André; PANJAN, Matjaž; FRANZ, Robert; ANDERSSON,
Joakim; NI, Pavel. Drifting potential humps in ionization
zones: the “propeller blades” of high power impulse magnetron
sputtering. Applied Physics Letters. 2013, vol. 103, no. 14, p. 144103.
38
BIBLIOGRAPHY
16. HECIMOVIC, Ante; BÖKE, Marc; WINTER, Jörg. The characteristic
shape of emission profiles of plasma spokes in HiPIMS: the
role of secondary electrons. Journal of Physics D: Applied Physics.
2014, vol. 47, no. 10, p. 102003.
17. BREILMANN, Wolfgang; EITRICH, A; MASZL, Cristian; HECIMOVIC,
Ante; LAYES, Vincent; BENEDIKT, Jan; VON KEUDELL,
Achim. High power impulse sputtering of chromium: correlation
between the energy distribution of chromium ions and spoke
formation. Journal of Physics D: Applied Physics. 2015, vol. 48, no.
29, p. 295202.
18. MISHRA, Anurag; KELLY, Peter; BRADLEY, James W. The evolution
of the plasma potential in a HiPIMS discharge and its
relationship to deposition rate. Plasma Sources Science and Technology.
2010, vol. 19, no. 4, p. 045014.
19. ČAPEK, Jiřı; HÁLA, Matej; ZABEIDA, Oleg; KLEMBERG-SAPIEHA,
Jolanta-Ewa; MARTINU, Ludvik. Deposition rate enhancement
in HiPIMS without compromising the ionized fraction
of the deposition flux. Journal of Physics D: Applied Physics.
2013, vol. 46, no. 20, p. 205205.
20. POOLCHARUANSIN, Phitsanu; ESTRIN, Francis Lockwood;
BRADLEY, James W. The use of segmented cathodes to determine
the spoke current density distribution in high power impulse
magnetron sputtering plasmas. Journal of Applied Physics.
2015, vol. 117, no. 16, p. 163304.
21. BRITUN, Nikolay; PALMUCCI, Maria; KONSTANTINIDIS, Stephanos;
SNYDERS, Rony. Particle visualization in high-power
impulse magnetron sputtering. I. 2D density mapping. Journal of
Applied Physics. 2015, vol. 117, no. 16, p. 163302.
22. BRITUN, Nikolay; PALMUCCI, Maria; KONSTANTINIDIS, Stephanos;
SNYDERS, Rony. Particle visualization in high-power
impulse magnetron sputtering. II. Absolute density dynamics.
Journal of Applied Physics. 2015, vol. 117, no. 16, p. 163303.
23. HOLSTEIN, To. Imprisonment of resonance radiation in gases.
Physical Review. 1947, vol. 72, no. 12, p. 1212.
39
BIBLIOGRAPHY
24. JOLLY, Jacques; TOUZEAU, Michel. Measurement of metastablestate
densities by self-absorption technique. Journal of Quantitative
Spectroscopy and Radiative Transfer. 1975, vol. 15, no. 9, pp. 863–
872.
25. SCHULZE, Martin; YANGUAS-GIL, Angel; VON KEUDELL,
Achim; AWAKOWICZ, Peter. A robust method to measure metastable
and resonant state densities from emission spectra in argon
and argon-diluted low pressure plasmas. Journal of Physics D:
Applied Physics. 2008, vol. 41, no. 6, p. 065206.
26. BOFFARD, John B; JUNG, RO; LIN, Chun C; WENDT, AE. Measurement
of metastable and resonance level densities in rare-gas
plasmas by optical emission spectroscopy. Plasma Sources Science
and Technology. 2009, vol. 18, no. 3, p. 035017.
27. NAVRÁTIL, Zdeněk; DOSOUDILOVÁ, Lenka; HNILICA, Jaroslav;
BOGDANOV, Todor. Optical diagnostics of a surface-wavesustained
neon plasma by collisional–radiative modelling and
a self-absorption method. Journal of Physics D: Applied Physics.
2013, vol. 46, no. 29, p. 295204.
28. KRAMIDA, Alexander; RALCHENKO, Yuri; READER, Joseph;
TEAM, NIST ASD. NIST Atomic Spectra Database (version 5.9).
2021. Available also from: https://physics.nist.gov/asd.
29. GAILLARD, M; BRITUN, Nikolay; KIM, Yong M; HAN, Jeon G.
Titanium density analysed by optical absorption and emission
spectroscopy in a dc magnetron discharge. Journal of Physics D:
Applied Physics. 2007, vol. 40, no. 3, p. 809.
30. SAMUELSSON, Mattias; LUNDIN, Daniel; JENSEN, Jens; RAADU,
Michael A; GUDMUNDSSON, Jon Tomas; HELMERSSON,
Ulf. On the film density using high power impulse magnetron
sputtering. Surface and Coatings Technology. 2010, vol. 205, no. 2,
pp. 591–596.
31. BARKER, Paul Michael; LEWIN, Erik; PATSCHEIDER, Jörg. Modified
high power impulse magnetron sputtering process for increased
deposition rate of titanium. Journal of Vacuum Science
& Technology A: Vacuum, Surfaces, and Films. 2013, vol. 31, no. 6,
p. 060604.
40
BIBLIOGRAPHY
32. ANTONIN, Olivier; TIRON, Vasile; COSTIN, Claudiu; POPA,
Gheorghe; MINEA, Tiberiu. On the HiPIMS benefits of multipulse
operating mode. Journal of Physics D: Applied Physics. 2014,
vol. 48, no. 1, p. 015202.
33. BARKER, Paul Michael; KONSTANTINIDIS, Stephanos; LEWIN,
Erik; BRITUN, Nikolay; PATSCHEIDER, Jörg. An investigation
of c-HiPIMS discharges during titanium deposition. Surface and
Coatings Technology. 2014, vol. 258, pp. 631–638.
34. SOUČEK, Pavel; HNILICA, Jaroslav; KLEIN, Peter; FEKETE,
Matej; VAŠINA, Petr. Microstructure of titanium coatings controlled
by pulse sequence in multipulse HiPIMS. Surface and
Coatings Technology. 2021, vol. 423, p. 127624.
35. RAADU, Michael A; AXNÄS, Ingvar; GUDMUNDSSON, Jon
Tomas; HUO, Chunqing; BRENNING, Nils. An ionization region
model for high-power impulse magnetron sputtering discharges.
Plasma Sources Science and Technology. 2011, vol. 20, no. 6,
p. 065007.
41
Author’s publications not discussed in this work
• HNILICA, Jaroslav;, KLEIN, Peter; VAŠINA, Petr; SNYDERS, Rony;
BRITUN, Nikolay. Dynamics of sputtered particles in multipulse
HiPIMS discharge. Plasma Sources Sci. Technol. 2023, 32, 045003.
• ŠLAPANSKÁ, Marta; KROKER, Michael; KLEIN, Peter; HNILICA,
Jaroslav; VAŠINA, Petr. Spatially resolved study of spokes in
reactive HiPIMS discharge. Plasma Sources Sci. Technol. 2022, 31,
055010.
• ŠLAPANSKÁ, Marta; KROKER, Michael; HNILICA, Jaroslav; KLEIN,
Peter; VAŠINA, Petr. Single-shot spatial-resolved optical emission
spectroscopy of argon and titanium species within the spoke.
J. Phys. D: Appl. Phys. 2022, 55, 035205.
• SOUČEK, Pavel; HNILICA, Jaroslav; KLEIN, Peter; FEKETE, Matej;
VAŠINA, Petr. Microstructure of titanium coatings controlled by
pulse sequence in multipulse HiPIMS. Surf. Coat. Technol. 2021,
423, 127624.
• KLEIN, Peter; HNILICA, Jaroslav; FEKETE, Matej; ŠLAPANSKÁ,
Marta; VAŠINA, Petr. Spoke behaviour in reactive HiPIMS. Plasma
Sources Sci. Technol. 2021, 30, 055016.
• BERNÁTOVÁ, Katarína; KLEIN, Peter; HNILICA, Jaroslav; VAŠINA,
Petr. Temporal studies of titanium ionised density fraction
in reactive HiPIMS with nitrogen admixture. Plasma Sources Sci.
Technol. 2021, 30, 125002.
• BRITUN, Nikolay; HNILICA, Jaroslav. Optical spectroscopy for
sputtering process characterization. J. Appl. Phys. 2020, 127, 211101.
• ŠLAPANSKÁ, Marta; HECIMOVIC, Ante; GUDMUNDSSON,
Jon Tomas; HNILICA, Jaroslav; BREILMANN, Wolfgang; VAŠINA,
Petr; VON KEUDELL, Achim; Study of the transition from selforganised
to homogeneous plasma distribution in chromium
HiPIMS discharge. J. Phys. D: Appl. Phys. 2020, 53, 155201.
42
Author’s publications not discussed in this work
• BERNÁTOVÁ, Katarína; FEKETE, Matej; KLEIN, Peter; HNILICA,
Jaroslav; VAŠINA, Petr. Ionisation fractions of sputtered titanium
species at target and substrate region in HiPIMS. Plasma Sources
Sci. Technol. 2020, 29, 055010.
• FEKETE, Matej; BERNÁTOVÁ, Katarína; KLEIN, Peter; HNILICA,
Jaroslav; VAŠINA, Petr. Influence of sputtered species ionisation
on the hysteresis behaviour of reactive HiPIMS with oxygen
admixture. Plasma Sources Sci. Technol. 2020, 29, 025027.
• FEKETE, Matej; BERNÁTOVÁ, Katarína; KLEIN, Peter; HNILICA,
Jaroslav; VAŠINA, Petr. Evolution of discharge parameters
and sputtered species ionization in reactive HiPIMS with oxygen,
nitrogen and acetylene. Plasma Sources Sci. Technol. 2019, 28,
025011.
• SOUČEK, Pavel; DANIEL, Josef; HNILICA, Jaroslav; BERNÁTOVÁ,
Katarína;, ZÁBRANSKÝ, Lukáš; BURŠÍKOVÁ, Vilma; STUPAVSKÁ,
Monika; VAŠINA, Petr. Superhard nanocomposite nc-TiC/
a-C:H coatings: The effect of HiPIMS on coating microstructure
and mechanical properties. Surf. Coat. Technol. 2017, 311, 257–267.
• FALTÝNEK, Jan; HNILICA, Jaroslav; KUDRLE, Vít. Electron density
in amplitude modulated microwave atmospheric plasma
jet as determined from microwave interferometry and emission
spectroscopy. Plasma Sources Sci. Technol. 2017, 26, 015010.
• VORÁČ, Jan; SYNEK, Petr; POTOČŇÁKOVÁ, Lucia; HNILICA,
Jaroslav; KUDRLE, Vít. Batch processing of overlapping molecular
spectra as a tool for spatio-temporal diagnostics of power
modulated microwave plasma jet. Plasma Sources Sci. Technol.
2017, 26, 025010.
• VORÁČ, Jan; POTOČŇÁKOVÁ, Lucia; SYNEK, Petr; HNILICA,
Jaroslav; KUDRLE, Vít. Gas mixing enhanced by power modulations
in atmospheric pressure microwave plasma jet. Plasma
Sources Sci. Technol. 2016, 26, 025010.
• SCHÄFER, Jan; HNILICA, Jaroslav; ŠPERKA, Jiří; QUADE, Antje;
KUDRLE, Vít; FOEST, Rüdiger; VODÁK, Jiří; ZAJÍČKOVÁ,
43
Author’s publications not discussed in this work
Lenka. Tetrakis(trimethylsilyloxy)silane for nanostructured SiO
2 -like films deposited by PECVD at atmospheric pressure. Surf.
Coat. Technol. 2016, 295, 112–118.
• HNILICA, Jaroslav; POTOČŇÁKOVÁ, Lucia; KUDRLE, Vít. Time
resolved optical emission spectroscopy in power modulated
atmospheric pressure plasma jet. Open Chemistry 2015, 13, 577–
585.
• POTOČŇÁKOVÁ, Lucia; HNILICA, Jaroslav; KUDRLE, Vít. Spatially
resolved spectroscopy of an atmospheric pressure microwave
plasma jet used for surface treatment. Open Chemistry 2015,
13, 541–548.
• VORÁČ, Jan; HNILICA, Jaroslav; KUDRLE, Vít; DVOŘÁK, Pavel;
Spatially resolved measurement of hydroxyl (OH) radical concentration
in a microwave plasma jet by planar laser-induced
fluorescence. Open Chemistry 2015, 13, 193–197.
• PAVLIŇÁK, David; HNILICA, Jaroslav; QUADE, Antje; SCHÄFER,
Jan; ALBERTI, Milan; KUDRLE, Vít. Functionalisation and pore
size control of electrospun PA6 nanofibres using a microwave
jet plasma. Polym. Degrad. Stab. 2014, 108, 48–55.
• HNILICA, Jaroslav; POTOČŇÁKOVÁ, Lucia; KUDRLE, Vít. Modulation-induced
filament tearing in microwave atmospheric plasma
jet. IEEE Trans. Plasma Sci. 2014, 42, 2472–2473.
• MANAKHOV, Anton; ZAJÍČKOVÁ, Lenka; ELIÁŠ, Marek; ČECHAL,
Jan; POLČÁK, Josef; HNILICA, Jaroslav; BITTNEROVÁ,
Š´těpánka; NEČAS, David. Optimization of cyclopropylamine
plasma polymerization toward enhanced layer stability in contact
with water. Plasma Process. Polym. 2014, 11, 532–544.
• HNILICA, Jaroslav; KUDRLE, Vít. Time-resolved study of amplitude
modulation effects in surface-wave atmospheric pressure
argon plasma jet. J. Phys. D: Appl. Phys. 2014, 47, 085204.
• HNILICA, Jaroslav; POTOČŇÁKOVÁ, Lucia; STUPAVSKÁ, Monika;
KUDRLE, Vít. Rapid surface treatment of polyamide 12 by microwave
plasma jet. Appl. Surf. Sci. 2014, 288, 251–257.
44
Author’s publications not discussed in this work
• POTOČŇÁKOVÁ, Lucia; HNILICA, Jaroslav; KUDRLE, Vít. Increase
of wettability of soft- and hardwoods using microwave
plasma. Int. J. Adhes. Adhes. 2013, 45, 125—131.
• NAVRÁTIL, Zdeněk; DOSOUDILOVÁ, Lenka; HNILICA, Jaroslav;
BOGDANOV, Todor. Optical diagnostics of a surface-wave-sustained
neon plasma by collisional–radiative modelling and a selfabsorption
method. J. Phys. D: Appl. Phys. 2013, 46, 295204.
• HNILICA, Jaroslav; SCHÄFER, Jan; FOEST, Rüdiger; ZAJÍČKOVÁ,
Lenka; KUDRLE, Vít. PECVD of nanostructured SiO 2 in a modulated
microwave plasma jet at atmospheric pressure. J. Phys. D:
Appl. Phys. 2013, 46, 335202.
• HNILICA, Jaroslav; KUDRLE, Vít; VAŠINA, Petr; SCHÄFER, Jan;
AUBRECHT, Vladimír. Characterization of a periodic instability
in filamentary surface wave discharge at atmospheric pressure
in argon. J. Phys. D: Appl. Phys. 2012, 45, 055201.
• HNILICA, Jaroslav; POTOČŇÁKOVÁ, Lucia; KUDRLE, Vít. Surface
treatment by atmospheric-pressure surfatron jet. IEEE Trans.
Plasma Sci. 2012, 40, 2925–2930.
• SCHÄFER, Jan; VAŠINA, Petr; HNILICA, Jaroslav; FOEST, Rudiger;
KUDRLE, Vít; WELTMANN, Klaus-Dieter. Visualization
of revolving modes in RF and MW nonthermal atmospheric
pressure plasma jets. IEEE Trans. Plasma Sci. 2011, 39, 2350–2351.
45