Martin P.M. Handbook of Deposition Technologies for Films and Coatings, Third Edition: Science, Applications and Technology
Подождите немного. Документ загружается.
Unfiltered and Filtered Cathodic Arc Deposition 511
beyond 50 V; the hardness increased up to 35 GPa. The coatings started to oxidize when the
temperature exceeded 1000
◦
C, which is slightly better than TiAlN.
The trend of adding more components, going even beyond quaternary compounds, has become
popular in recent years because coatings can be tailored to the applications. In general,
combinations of the materials Ti–Al–Cr–Si–C–B–N are investigated, and other transition
metals are also considered, such as Nb, V, Mo, Ta, and Hf.
10.6.1.4 Multilayers and Nanolaminates
A relatively straightforward approach to a (nano)structured coating is to use the multilayer
deposition approach. Multilayer structures often outperform monolithic layers for a variety of
reasons such as impeding dislocation movement and crack propagation. The total thickness of
multilayers is often greater than 1 m, while individual layers are just a few nanometers or
tens of nanometers. When the individual layers of a multilayer system have ∼10 nm or less
thickness, one speaks of nanolayers. There are many papers devoted to the subject, and only
some brief comments and literature references can be given here [197–210].
In most cases, the individual layers have a distinct composition and microstructure and an
interface or buffer layer can be used to improve adhesion to the substrate. It is also possible to
produce multilayers from only one element by periodically changing the microstructure via
the deposition conditions. One example of the latter is a carbon–carbon structure where the
layers have different sp
3
to sp
2
ratios [211, 212] (more about amorphous carbon near the end
of this chapter).
One reason for using multilayers is that one can influence hardness and stress in a relatively
decoupled way. For example, an optimized nanolayered WC–CrAlN heterostructure on silicon
and S45C steel with a bilayer period of 2–10 nm exhibited a residual stress of less than 2 GPa,
while the microhardness was very high, in the range 30–43 GPa [213].
When considering a nanostructure it is clear that the presence of macroparticles is detrimental
because macroparticles are usually larger than the thickness of the individual layer or
characteristic size of the grain, and sometimes they even exceed the thickness of the whole
coating. To make matters worse, as mentioned before, the defects introduced by a
macroparticles are larger than the macroparticles themselves. Therefore, it is highly desirable
to improve the coating by eliminating macroparticles, which is done by switching to filtered
arc deposition (accepting the loss of high rates; Section 10.7) or to a different technique
altogether (losing some advantages of the self-ion-assisted process).
10.6.2 Decorative Coatings
Most coatings are actually multifunctional coatings. For example, the much-discussed TiN is
used for its hardness, wear- and corrosion-resistant properties, as well as for its golden color.
512 Chapter 10
The appearance of a coating is the result of extrinsic and intrinsic optical properties, where
extrinsic refers to alteration of the light direction, resulting in matt or glossy surfaces, and
intrinsic refers to wavelength-dependent (spectrally selective) reflection and absorption of the
incident light.
The color of the coating results from the reflected light and therefore we need to consider
interference of multiply reflected waves (in the case of transparent or slightly absorbing film
materials), and the absorbing properties for spectrally absorbing coatings. The classic
examples of decorative coatings, TiN and ZrN, fall in the latter category and deserve closer
examination. Spectrally selective reflection is closely related to the incident light being
spectrally selectively absorbed.
The most important process for arc-deposited decorative coatings like TiN is free electron
absorption or intraband transition of electrons from partially filled valence bands to unfilled
states just above the Fermi level. The electrons can absorb a broad range of photon energies,
including photons in the visible part of the spectrum. The quantum-mechanical selection rules
for electron transitions can force electrons to return to lower levels via different levels than
their excitation paths, and the color of emitted light may not correspond to the absorbed light,
contributing to color appearance.
While intraband transitions apply to conducting materials, the color of semiconducting or
dielectric materials is mostly based on interband transitions. They occur when there is a filled
valence band and an empty conduction band, separated by an energy gap. Photons with energy
greater than the gap will cause electron transitions from the valence band into the conduction
band, and the color results from the remaining, not absorbed, visible light. Materials whose
bandgap is wider than the ultraviolet limit of the visible spectrum, E
g
> 3.5 eV, are transparent
to the eye but color can appear by interference of reflected light. Narrow bandgap materials,
whose bandgap is smaller than the infrared limit of the visible spectrum, E
g
< 1.7 eV, appear
black since all of the visible light photons are sufficiently energetic to excite electrons into the
conduction band. Material in the intermediate region, with a bandgap of 1.7 eV < E
g
< 3.5 eV,
can show a characteristic color.
Color can also be caused by defects, e.g. in form of impurities that occupy an energetic state in
the energy gap. Valence electrons can be excited into these states, giving rise to spectrally
selective absorption, or color. Examples of this mechanism are the coloration of corundum
(Al
2
O
3
) and diamond (C) by doping.
The most relevant arc-deposited decorative coating is TiN, and so let us have a closer look at
the deposition–property relationship for this material. The golden color is mainly due to
intraband transitions of the free electrons [214, 215].
Unfiltered and Filtered Cathodic Arc Deposition 513
In a simplified picture, the stoichiometry of the nitride determines the electrical and optical
properties: the greater the nitrogen content, the lower the density of free electrons. Similar to
having a plasma frequency defined for electrons in the plasma, one may consider
electrons in the conduction band as free (Drude model) and assign them a plasma frequency,
too,
ω
pl
=
n
e
e
2
ε
0
m
e
1/2
(10.43)
where m
e
stands for the effective electron mass, which is not exactly the same as the free
electron mass we use in the plasma because here the electron–lattice interaction is taken into
account in this way. The complex refractive index
N = n + ik = 1 −
ω
2
pl
ω
2
(10.44)
is zero at the plasma frequency, where k is the extinction coefficient and n is the real part of the
refractive index. The plasma frequency separates the frequency domains that are either
reflecting and absorbing ( <<
pl
) or transparent ( >>
pl
) for electromagnetic waves of
frequency ω. The region around the plasma frequency is also called the plasma edge.
Because the free carrier density of solids like TiN is rather high, typically about 10
28
m
−3
, the
corresponding plasma frequency is about 5 ×10
15
s
−1
, which corresponds to waves in the
ultraviolet (UV) region. Therefore, metals are reflecting and absorbing for all wavelengths
longer than the UV spectral region, including the visible. For the conducting nitrides, the
plasma frequency is reduced as the nitrogen content is increased. This leads to a more yellow
appearance with an overall reduced reflectivity [215].
The same trends also hold for ZrN. As the nitrogen content increases, the plasma edge and
related reflectivity minimum in the UV shift to longer wavelengths. The color of ZrN is more
sensitive to changes in the composition than TiN. For the latter, changing the N/Ti ratio from
under-stoichiometric 0.92 to over-stoichiometric 1.14 is hardly noticeable to the eye.
While TiN (gold) and TiC (dark gray) have a limited range of color, addition of another metal
can have a dramatic effect, such as producing dark blue (Ti,Al)N. To obtain gray to black one
may consider TiAlN with high Al content, while low Al content leads to a violet–bronze
appearance. ZrCN exhibits a wide range of colors from stainless steel and nickel to brass, gold,
and black [216]. The colors of TiCN coatings range from copper–reddish (bronze) to
blue–gray. Ternary nitrides such (Ti,V)N, (Zr,V)N, (Zr,Al)N, and (Zr,Cr)N have an even
greater range of possible colors [217].
514 Chapter 10
10.7 Deposition of Coatings Using Filtered Cathodic Arc Plasmas
10.7.1 Tetrahedral Amorphous Carbon
Cathodic arc deposition is a technique that can produce diamond-like amorphous carbon films
with very high sp
3
content, such that they are sometimes called amorphous diamond films
[218–220]. For most applications of such films, the incorporation of macroparticles is not
acceptable and therefore the case of a-C and ta-C films is considered here, under the heading
filtered arc deposition. The now common term ta-C stands for tetrahedral amorphous carbon,
indicating the dominance of tetrahedral or sp
3
hybridized carbon atoms. In contrast to many
other deposition techniques, filtered cathodic arcs can operate in a vacuum, using a graphite
cathode; the resulting films are hydrogen free apart from a small contamination due to the
residual water in the vacuum chamber. Other elements, including hydrogen [221] or nitrogen
[222–226], can be added when operating in a background gas containing hydrocarbons or
ammonia, for example, or one can use metal organics or metal plasma from another discharge
[227, 228].
Much has been published since the massive research efforts of the 1990s, and so it must suffice
to summarize a few selected results and to point to some excellent reviews (e.g. [229–231]).
The earliest ta-C films were deposited in the late 1970s, and in fact ta-C films drove the
development of the early macroparticle filters [232–234]. Much research followed in the 1990s
when the extraordinary mechanical, chemical, optical, and electronic properties of a-C and ta-C
were fully recognized [225, 235–240] and applications in the information storage [241–244],
automotive/tooling [203, 245–247], and biomedical [248–250] industries were evident.
Using graphite as the feedstock material (cathode), carbon plasma is produced that contains
mostly singly charged ions with a most likely kinetic energy of about 20 eV (see Table 10.2).
The properties of the diamond-like material depend largely on the acceleration that carbon
ions experience in the sheath to the substrate, which can be controlled via substrate bias [239]
or plasma bias [251]. Deposition is generally done at room temperature because high
temperature degrades the diamond-like properties of the material by reducing the density of
metastable sp
3
bonds [252–254]. While the sp
3
hybrids determine the diamond-like properties
such as high hardness, high Young’s modulus, chemical inertness, and a wide optical band gap
[255], the sp
2
hybrids are largely responsible for the electronic and optical properties because
the π states lie closest to the Fermi level [256].
For applications in the magnetic storage industry, ultrathin a-C films are wanted because they
combine corrosion protection with advantageous mechanical properties, and not least, because
they allow wetting with a lubricant. In order to bring the read–write head as close as possible
to the magnetic layer containing the magnetically stored information, the films on the medium
(like a hard disk) and read–write head must be thin as possible, but continuous. The industry
Unfiltered and Filtered Cathodic Arc Deposition 515
has succeeded in producing films as thin as 2 nm yet meeting the tough quality requirements,
and efforts are underway to reduce the thickness even further. It is clear that we quickly run
into the fundamental limits of the possible.
At the other extreme, where thick (> 1 m) a-C or ta-C films are wanted, e.g. for
microelectromechanical systems (MEMS) fabrication, the reduction of residual stress is
paramount to prevent catastrophic failure by delamination. Repeated rapid thermal annealing
up to 600
◦
C can be used to remove most of the compressive stress from ta-C films [252].
Another approach is the previously described bombardment with energetic ions (Section
10.5.2) [257, 258].
10.7.2 High-Quality Metal Films
As the expectations and requirements on coatings increase and other techniques deliver very
good coatings, cathodic arc deposition needs to improve too, and that means addressing the
macroparticle issue. The most complete and successful approach to date is the incorporation of
a macroparticle filter, as mentioned in Section 10.4.3. This, however, means accepting plasma
losses and associated reductions in the deposition rate, which is a major reason why filtered arc
deposition is not widely implemented. Filtered arc deposition is used when dealing with high
value added components.
An example of advanced filtered arc coatings is the precision metallization of semiconductors
and MEMS structures. Using optimized pulsed bias, angle-dependent self-sputtering can be
utilized to deal with even very small three-dimensional structures [259–261]. One of the
impressive images is the near-conformal coating of a 1 m trench structure with tantalum
using a high-current pulsed filtered arc (Figure 10.21). More recently, even smaller structures
have been successfully filled with copper using a filtered arc [262].
Figure 10.21: Conformal barrier coating of tantalum on a lithographically etched structure.
(Photograph courtesy of P. Siemroth, Germany.)
516 Chapter 10
10.7.3 Optical Films
Macroparticle contamination is especially undesirable for optical coatings and coatings related
to optoelectronic devices because they deteriorate the appearance, or worse, may make the
device dysfunctional. While such films can be made with other deposition techniques such as
reactive magnetron sputtering or chemical vapor deposition (CVD), the self-ion-assistance of
cathodic arc deposition gives the opportunity to produce superior coatings. For example, the
refractive indices of optically transparent oxides are higher than those of their sputtered
counterparts owing to the reduction or elimination of intercolumnar voids [263–265], and
thereby they are also less prone to changes that occur owing to the uptake of moisture.
Oxide coatings can be readily obtained with the filtered cathodic arc operating in a
low-pressure background of oxygen. The oxygen partial pressure is typically in the range of
several Pa and must be adjusted depending on the arrival rate of metal on the substrate surface.
Pulsed systems with lower average deposition rates can produce stoichiometric oxide films at
even lower pressure.
The arc deposition process is relatively forgiving to variations in the oxygen partial pressure:
the arc will operate in a stable manner and the films tend to be fully stoichiometric. In contrast
to magnetron sputtering, the promotion of arcing by cathode poisoning is not an issue because,
after all, we want to use the arc process in the cathodic arc anyway. Examples of high-quality
optical films include Al
2
O
3
,TiO
2
,VO
2
,Nb
2
O
5
, and ZrO
2
, with typical deposition rates of
25–50 nm/minute [266]. Despite ion assistance in the growth, the substrate temperature
remains an important parameter that affects phase and texture of the coating.
Of special interest and importance is TiO
2
because it can have different phases: amorphous,
brookite, anatase, and rutile. Cathodic arc deposition offers the possibility of synthesizing the
rutile phase at room temperature provided the ions are accelerated by biasing to about 100 eV
or greater [263, 267]. Lower bias favors the anatase phase, which is very important for
photocatalytic applications.
Another material of interest is MgO, which is used as a transparent protective layer in
alternating current plasma displays because it is highly transparent and has a high packing
density and surface binding energy leading to good sputter resistance [268].
Cathodic-arc-deposited MgO films grew preferably with (200) and (220) orientation on glass
and silicon when the partial pressures was increased from 1 to 4 Pa [269]. With increasing
oxygen flow, the grain size increased from typically about 10 to about 25 nm, and the slight
absorption seen in the blue and green disappeared: the transmittance exceeded 95% throughout
the visible spectrum for the 1.0–1.3 m thick films [269].
Yet another oxide synthesized by filtered cathodic arcs is aluminum oxide, which is interesting
because it combines low absorption with high Young’s modulus and hardness [270, 271]. The
Unfiltered and Filtered Cathodic Arc Deposition 517
refractive index increased markedly when the substrate temperature was varied from 200
◦
Cto
300
◦
C, indicating that ion assistance does not fully replace substrate heating, i.e. the substrate
temperature remains an important parameter in energetic deposition.
The transparent optical materials are not limited to oxides; there are a number of wide bandgap
nitrides such as AlN. M
¨
andl et al. [272] produced AlN at substrate temperatures from about
room temperature to 370
◦
C using continuous arc operation with a high deposition rate of
85–115 nm/minute. While thin films of 30 nm were amorphous, much thicker films, 1–3 m,
showed the NaCl structure with a microhardness of 18–20 GPa. As with most optical
materials, AlN films are highly insulating with a resistivity of 1–2 ×10
9
m. Besides being an
interesting optical material, AlN films of thickness 1–3 m are also good oxidation protection
coatings up to 1100
◦
C [273].
10.7.4 Transparent Conducting Oxides
A special class of transparent oxide films comprises those that are electrically conducting.
While most oxides are good insulators, some are wide bandgap semiconductors, usually
of n-type, which have gained enormous economic importance in recent years owing to
their use in numerous devices such as flat panel displays, solar cells, and touch screens.
Indium–tin oxide (ITO) is the most common transparent conducting oxide (TCO) because it
combines very high transmittance (> 80%) with very low electrical resistivity
(∼10
4
cm=1m). While ITO can be deposited by cathodic arc deposition [274], much
research aims to find alternative, well-performing yet lower cost materials because the price
of indium has become very high. Cathodic arc deposition plays an important role in this effort;
in many cases, doped SnO
2
and ZnO films are being investigated [275]. Besides this
cost-driven effort, research is also underway to find well-performing p-type TCOs
[276, 277].
Pioneering work was done by Boxman and Goldsmith’s group at Tel Aviv University [278].
Using a 90
◦
filtered arc system they deposited SnO
2
on various substrates using a tin (Sn)
cathode operating in an oxygen background. An optimum window of pressure was identified,
with the best material made at 6 mtorr (0.8 Pa). The lower limit of the process window (3 mtorr
or 0.4 Pa) was determined by the lack of oxygen to obtain the stoichiometric SnO
2
, and the
upper limit (12 mtorr or 1.6 Pa) was given by a drastically reduced deposition rate. The
as-deposited films usually appear brownish or yellowish. Rapid thermal annealing at 300
◦
Cin
air reduces the resistivity by about one order of magnitude and the films become clear with a
transmittance of about 85% in the visible.
Many more research projects have followed [275, 276, 279–289] as reviewed by Tay [288] and
in a recent book [3].
518 Chapter 10
10.7.5 Nitride Multilayers, Nanocomposites, and MAX Phases
Large area coatings with good uniformity are a challenge for filtered cathodic arcs. Those
issues can be solved, as was demonstrated by Gorokhovsky et al., who developed their
dual-arc large-area filtered arc deposition (LAFAD) facility [290], a commercial-size batch
coater for high value added products. They deposited TiN/Ti multilayers and TiCrN/TiCr
superlattice multilayer coatings on periodontal instruments [200].
As mentioned in Section 10.6.1, multilayers and nanostructures are of increasing relevance to
modern applications. It is common to require that advanced structures do not suffer from
macroparticle- or macroparticle-induced defects; hence, filtered arc deposition is considered
here. This is especially true for an emerging class of materials called MAX phases [291],
which consist of an early transition metal M, an element from the A groups of the Periodic
Table, usually IIIA and IVA, and a third element, X, which is either nitrogen or carbon. The
composition is M
n+1
AX
n
, where n = 1, 2 or 3. MAX phases form three groups, based on the
number of atoms of the M, A, and X elements in each unit cell; these groups are known as 211,
312, and 413 materials. Examples for the first group are M
2
AlC, with M = Sc, Y, La, Ti, Zr, Hf,
V, Nb, Ta, Cr, Mo, W. More than 50 MAX phases are known today and more can be expected
since this is a field of intense research. MAX phase materials have amazing properties because
they combine the high temperature stability of ceramics with properties of metals such as good
thermal and electrical conductivity. They are machinable, in contrast to conventional ceramics.
Figure 10.22: Cross-sectional TEM image of a Ti
2
AlC MAX-phase coating synthesized by filtered
three-cathode pulsed arc deposition at 900
C: (a) low-resolution image, with the α–Al
2
O
3
(0001)
substrate at the bottom, coating in center, and the amorphous-looking TEM-glue layer on top;
(b) high-resolution image clearly showing the nanolaminate structure. (Image courtesy of J. Ros
´
en
(deposition) and P. Persson (TEM), work done at Sydney University.)
Unfiltered and Filtered Cathodic Arc Deposition 519
The synthesis of the Ti
2
AlC MAX phase was successfully demonstrated using filtered
cathodic arc deposition with a triple-cathode pulsed arc system [292]. The Ti:Al:C pulse ratios
were fine-tuned to obtain the necessary stoichiometric relation in the coating, which was
deposited on α−Al
2
O
3
(0001) substrates at 900
◦
C. Transmission electron microscopy (TEM)
revealed epitaxial growth and the nanolaminated hexagonal structure with some rare cubic
precipitates (Figure 10.22).
There are many more applications of unfiltered and filtered cathodic arc deposition, and the
reader is referred to a recent, more comprehensive overview [3].
References
[1] J. Priestley, Experiments in which rings, consisting of all the prismatic colours, were made by electrical
explosions on the surface of metals, in: The History and Present State of Electricity, Vol. II, London (1775)
329–335.
[2] J. Priestley, The History and Present State of Electricity, 3rd ed, London (1775).
[3] A. Anders, Cathodic Arcs: From Fractal Spots to Energetic Condensation, Springer, New York (2008).
[4] S. Anders, A. Anders, On modes of arc cathode operation, IEEE Trans. Plasma Sci. 19 (1991) 20–24.
[5] A. Anders, B. J
¨
uttner, Cathode mode transition in high-pressure discharge lamps at start-up, Lighting Res.
Technol. (GB) 22 (1990) 111–115.
[6] H. Ehrich, B. Hasse, M. Mausbach, K.G. M
¨
uller, The anodic vacuum arc and its application to coatings,
J. Vac. Sci. Technol. A 8 (1990) 2160–2164.
[7] I.I. Beilis, R.L. Boxman, S. Goldsmith, V.L. Paperny, Radially expanding plasma parameters in a hot
refractory anode vacuum arc, J. Appl. Phys. 88 (2000) 6224–6231.
[8] R.H. Fowler, L. Nordheim, Electron emission in intense electric fields, Proc. R. Soc. (London) A119 (1928)
173–181.
[9] E. Hantzsche, The thermo-field emission of electrons in arc discharges, Beitr. Plasmaphys. 22 (1982)
325–346.
[10] G.A. Mesyats, D.I. Proskurovsky, Pulsed Electrical Discharge in Vacuum, Springer, Berlin (1989).
[11] W.W. Dolan, W.P. Dyke, Temperature and field emission of electrons from metals, Phys. Rev. 35 (1954)
327.
[12] E.L. Murphy, R.H. Good Jr., Thermionic emission, field emission, and the transition region, Phys. Rev. 102
(1956) 1446.
[13] G.A. Mesyats, Explosive Electron Emission, URO Press, Ekaterinburg (1998).
[14] S.P. Bugaev, E.A. Litvinov, G.A. Mesyats, D.I. Proskurovskii, Explosive emission of electrons, Sov. Phys.
Usp. 18 (1975) 51–61.
[15] E.A. Litvinov, G.A. Mesyats, D.I. Proskurovskii, Field emission and explosive emission processes in
vacuum discharges, Sov. Phys. Usp. 26 (1983) 138.
[16] W.G. Chace, H.K. More, Exploding Wires, Plenum Press, New York (1962).
[17] Y.A. Kotov, V.S. Sedoi, L.I. Chemezova, The integral of action and the energy for electrically exploded
wires, Institute of High Current Electronics, Tomsk (1986).
[18] R.P.P. Smeets, The origin of current chopping in vacuum arcs, IEEE Trans. Plasma Sci. 17 (1989) 303–310.
[19] A. Anders, Chopping effect observed at cathodic arc initiation, IEEE Trans. Plasma Sci. 28 (2000)
1303–1304.
[20] G.A. Mesyats, Cathode Phenomena in a Vacuum Discharge: The Breakdown, the Spark and the Arc, Nauka,
Moscow (2000).
[21] M.B. Bochkarev, Ecton processes of low current vacuum arc imaged with streak technique, in: XXIth Int.
Symp. Discharges and Electrical Insulation in Vacuum, Yalta, Ukraine (2004) 241–244.
520 Chapter 10
[22] J. Prock, Solidification of hot craters on the cathode of vacuum arcs, J. Phys. D: Appl. Phys. 19 (1986)
1917–1924.
[23] A. Anders, S. Anders, A. F
¨
orster, I.G. Brown, Pressure ionization: its role in metal vapor vacuum arc
plasmas and ion sources, Plasma Sources Sci. Technol. 1 (1992) 263–270.
[24] J.E. Daalder, Random walk of cathode arc spots in vacuum, J. Phys. D: Appl. Phys. 16 (1983) 17–27.
[25] E. Hantzsche, B. J
¨
uttner, H. Pursch, On the random walk of arc cathode spots in vacuum, J. Phys. D: Appl.
Phys. 16 (1983) L173–L179.
[26] B. J
¨
uttner, V.F. Puchkarev, E. Hantzsche, I. Beilis, Cathode spots, in: R.L. Boxman, D.M. Sanders, P.J.
Martin (Eds.), Handbook of Vacuum Arc Science and Technology, Noyes, Park Ridge, New Jersey (1995)
73–281.
[27] B. J
¨
uttner, The dynamics of arc cathode spots in vacuum: new measurements, J. Phys. D: Appl. Phys. 30
(1997) 221–229.
[28] B. J
¨
uttner, H. Pursch, V.A. Shilov, The influence of surface roughness and surface temperature on arc spot
movement in vacuum, J. Phys. D: Appl. Phys. 17 (1984) L31–L34.
[29] I. Beilis, B.E. Djakov, B. J
¨
uttner, H. Pursch, Structure and dynamics of high-current arc cathode spots in
vacuum, J. Phys. D: Appl. Phys 30 (1997) 119–130.
[30] B. J
¨
uttner, I. Kleberg, The retrograde motion of arc cathode spots in vacuum, J. Phys. D: Appl. Phys. 33
(2000) 2025–2036.
[31] A. Anders, The fractal nature of cathode spots, IEEE Trans. Plasma Sci. 33 (2005) 1456–1464.
[32] M. Schroeder, M. Fractals, M. Chaos, Power Laws: Minutes from an Infinite Paradise, 8th ed., W.H.
Freeman and Co, New York (2000).
[33] I. Kleberg, Dynamics of cathode spots in external magnetic field, Humboldt University, Berlin (2001) (in
German).
[34] A.P. Ehiasarian, P.E. Hovsepian, R. New, J. Valter, Influence of steering magnetic field on the time-resolved
plasma chemistry in cathodic arc discharges, J. Phys. D: Appl. Phys. 37 (2004) 2101–2106.
[35] K.K. Zabello, Y.A. Barinov, A.M. Chaly, A.A. Logatchev, S.M. Shkol’nik, Experimental study of cathode
spot motion and burning voltage of low-current vacuum arc in magnetic field, IEEE Trans. Plasma Sci. 33
(2005) 1553–1559.
[36] D.Y. Fang, Temperature dependence of retrograde velocity of vacuum arcs in magnetic fields, IEEE Trans.
Plasma Sci. 11 (1983) 110–114.
[37] A.E. Robson, A. von Engel, Origin of retrograde motion of arc cathode spots, Phys. Rev. 93 (1954)
1121–1122.
[38] K.K. Zabello, Y.A. Barinov, A.A. Logatchev, S.M. Shkol’nik, Cathode spot motion and burning voltage of
low-current vacuum arc with electrodes of copper–chromium composition in magnetic field, in: Proc. of
XXIth Int. Symp. Discharges and Electrical Insulation in Vacuum, Yalta, Ukraine (2004) 280–281.
[39] B.B. Mandelbrot, The Fractal Geometry of Nature, W.H. Freeman and Co., New York (1983).
[40] R.P.P. Smeets, F.J.H. Schulpen, Fluctuations of charged particle and light emission in vacuum arcs, J. Phys.
D: Appl. Phys. 21 (1988) 301–310.
[41] A. Anders, E.M.Yu. Oks, G. Yushkov, Cathodic arcs: fractal voltage and cohesive energy rule, Appl. Phys.
Lett. 86 (2005) 211503–1–3.
[42] S. Anders, B. J
¨
uttner, Influence of residual gases on cathode spot behavior, IEEE Trans. Plasma Sci. 19
(1991) 705–712.
[43] T. Sch
¨
ulke, P. Siemroth, Vacuum arcs cathode spots as a self-similarity phenomenon, IEEE Trans. Plasma
Sci. 24 (1996) 63–64.
[44] P. Siemroth, T. Sch
¨
ulke, T. Witke, Investigations of cathode spots and plasma formation of vacuum arcs by
high speed microscopy and spectrography, IEEE Trans. Plasma Sci. 25 (1997) 571–579.
[45] E. Hantzsche, Mysteries of the arc cathode spot: a retrospective glance, IEEE Trans. Plasma Sci. 31 (2003)
799–808.
[46] A. Anders, G.Y. Yushkov, Angularly resolved measurements of the ion energy of vacuum arc plasmas, Appl.
Phys. Lett. 80 (2002) 2457–2459.