Coatings Technology Handbook 2010 Part 5 ppsx

26

-1

26

Electrodeposition

of Polymers

26.1 Introduction

26-

1
26.2 Advantages

26-

1
26.3 History

26-

2
26.4 Process

26-

2

26.5 Equipment

26-

3

26.6 Laboratory 5
References

26-

5
Bibliography

26-

6

26.1 Introduction

The electrodeposition of polymers is an extension of painting techniques into the field of plating and,
like plating, is a dip coating process. The art of metal plating utilizes the fact that metal ions, usually
Ni

2+

or Cu

2+

, can be discharged on the cathode to give well-adhering deposits of metallic nickel, copper,
etc. The chemical process of deposition can be described as 1/2 Me

2+



+

1

F

(or 96,500 coulombs) of
electrons gives 1/2 Me

0

. In the case of electrodeposition of ionizable polymers, the deposition reaction
is described as R

3

NH

+

OH

–



+

1

F



→

R

3

N

+

H

2

O or the conversion of water-dispersed, ammonium-type
ions into ammonia-type, water-insoluble polymers known as cathodic deposition. Alternatively, a large
number of installations utilize the anodic deposition process RCOO

–



+

H

+

less 1

F



→

RCOOH. It should
be mentioned that “R” symbolizes any of the widely used polymers (acrylics, epoxies, alkyds, etc.).
The electrodeposition process is defined as the utilization of “synthetic, water dispersed, electrodepos-
itable macro-ions.”

1

26.2 Advantages

Metal ions, typically 1/2 Ni

2+

, show an electrical equivalent weight 1/2 Ni

2+

equal to approximately 29.5
g, while the polymeric ions typically used for electrodeposition exhibit a gram equivalent weight (GEW)
of approximately 1600. Thus, 1

F

plates out of 30 g of nickel and deposits 1600 g of macroions. If we

George E. F. Brewer*

George E. F. Brewer Coating
Consultants

* Deceased.

DK4036_book.fm Page 1 Monday, April 25, 2005 12:18 PM
© 2006 by Taylor & Francis Group, LLC
Throwing Power • Maintaining a Steady State • Rupture Voltage
Conveyors • Metal Preparation • Tank Enclosures • Dip Tanks •
Wate r • Bake or Cure
Rectifiers • Counterelectrodes • Agitation • Temperature
Control • Ultrafilter • Paint Filters • Paint Makeup • Deionized

27

-1

27

Electroless Plating

27.1 Introduction

27-

1
27.2 Plating Systems

27-

2
27.3 Electroless Plating Solutions

27-

3

27.4 Practical Applications

27-

4
27.5

27.6 Stability of Plating Solutions

27-

7
27.7 Electroless Plating

27-

7

27.8 Properties of Chemically Deposited Metal Coatings

27-

10
References

27-

11

27.1 Introduction

In electroless plating, metallic coatings are formed as a result of a chemical reaction between the reducing
agent present in the solution and metal ions. The metallic phase that appears in such reactions may be
obtained either in the bulk of the solution or as a precipitate in the form of a film on a solid surface.
Localization of the chemical process on a particular surface requires that the surface must serve as a
catalyst. If the catalyst is a reduction product (metal) itself, autocatalysis is ensured, and in this case, it
is possible to deposit a coating, in principle, of unlimited thickness. Such autocatalytic reactions constitute

the essence of practical processes of electroless plating. For this reason, these plating processes are
sometimes called autocatalytic.
Electroless plating may include metal plating techniques in which the metal is obtained as a result of
the decomposition reaction of a particular compound; for example, aluminum coatings are deposited
during decomposition of complex aluminum hydrides in organic solvents. However, such methods are
rare, and their practical significance is not great.
In a wider sense, electroless plating also includes other metal deposition processes from solutions in
which an external electrical current is not used, such as immersion, and contact plating methods in which
another more negative (active) metal is used as a reducing agent. However, such methods have a limited
application; they are not suitable for metallization of dielectric materials, and the reactions taking place
are not catalytic. Therefore, they usually are not classified as electroless plating.
Electroless plating now is widely used in modifying the surface of various materials, such as noncon-
ductors, semiconductors, and metals. Among the methods of applying metallic coatings, it is exceeded
in volume only by electroplating techniques, and it is almost equal to vacuum metallization.
Electroless plating methods have some advantages over similar electrochemical methods. These are as
follows:

A. Vakelis

Lithuanian Academy of Sciences

DK4036_book.fm Page 1 Monday, April 25, 2005 12:18 PM
© 2006 by Taylor & Francis Group, LLC
Deposition Rate • Solution Life • Reducing Agent Efficiency
Copper Deposition • Nickel Plating • Cobalt, Iron, and Tin
Factor • Solution Sensitivity to Activation
Plating • Deposition of Precious Metals • Deposition of Metal
Mechanisms of Autocatalytic Metal Ion Reduction
27-5
Alloys

27

-6

Coatings Technology Handbook, Third Edition

A more versatile explanation of the causes of catalysis in these processes is based on electrochemical
reactions. It is suggested that reducing agents are anodically oxidized on the catalyst surface and the
electrons obtained are transferred to metal ions, which are cathodically reduced. The catalytic process
comprises two simultaneous and mutually compensating electrochemical reactions. In this explanation
of the catalytic process, electrons are the active intermediate product. However, electrons are fundamen-
tally different from the conversational intermediate products of reactions. They may be easily transferred
along the catalyst without transfer of the mass, and for this reason, the catalyst reaction, contrary to all
other possible mechanisms (which are conventionally called “chemical mechanisms”), occurs not as a
result of direct contact between the reactants, or the reactants, or the reactant and an intermediate
substance, but because of the exchange of “anonymous” electrons via metal.
On the metal surface, when anodic oxidation of the reducer
(27.2)
and cathodic reduction of metal ions
(27.3)
proceed simultaneously, a steady state in the catalytic system of electroless plating is obtained, in which
the rates of both electrochemical reactions are equal, while the metal catalyst acquires a mixed potential

E

m

. The magnitude of this potential is between the equilibrium potentials

E

c

of the reducer and of the
metal. The specific value

E

m

depends on the kinetic parameters of these two electrochemical reactions.
Electrochemical studies of catalytic metal deposition reactions have shown that the electrochemical
mechanism is realized practically in all the systems of electroless plating.

4,6,7

At the same time, it has become clear that the process is often not so simple. It appears that anodic
and cathodic reactions occurring simultaneously often do not remain kinetically independent but affect
each other. For example, copper ion reduction increases along with anodic oxidation of formaldehyde.

8

The cathodic reduction of nickel ions and the anodic oxidation of hypophosphite in electroless nickel
plating solutions are faster than in the case in which these electrochemical reactions occur separately.
This interaction of electrochemical reactions probably is related to the changes in the state of the
metal–catalyst surface.
Electrochemical reactions may also hinder each other: for example, in reducing silver ions by hydrazine
from cyanide solutions, their rate is lower than is separate Ag–Ag(1) and redox systems.
The electrochemical nature of most of the autocatalytic processes discussed enables us to apply

electrochemical methods to their investigation. But, they must be applied to the entire system of electroless
plating, without separating the anodic and cathodic processes in space. One suitable method is based on
the measurement of polarization resistance. It can provide information on the mechanism of the process
and may be used for measuring the metal deposition rate (both in laboratory and in industry).

9

The
polarization resistance

R

p is inversely proportional to the process rate

i

:
(27.4)
(27.5)
where

b

a

and

b

c

are Tafel equation coefficients (b

≈

1/

α

nf

),

α

is the transfer coefficient,

n

is the number
of electrons taking part in the reaction for one molecule of reactant, and

f



=

F

/

RT

(

F



=

Faraday number).
Red →+Ox ne
Me
n+
+ ne
i
bb
Rb b
=
+
ac
pa c
()
R
dE
di

i
p
=






=0

DK4036_book.fm Page 6 Monday, April 25, 2005 12:18 PM
© 2006 by Taylor & Francis Group, LLC
27-12 Coatings Technology Handbook, Third Edition
13. G. Gawrilov, ChemischelStromläse/Vernickelung. Saulgau, Wurt.: Eugen Leuze Verlag, 1974.
14. K. M. Gorbunova et al., Fiziko-Khimichesklye Osnovy Processa Khimicheskogo Kobaltirovaniya.
Moscow: Nauka, 1974.
15. A. Molenaar and J. J. C. Coumans, Surface Technol., 16, 265 (1982).
16. Y. Okinaka, in Gold Plating Technology, H. Reid and W. Goldie, Eds. Hatch End, Middlesex, England:
Electrochemical Publications, Ltd., 1974, p. 82.
DK4036_book.fm Page 12 Monday, April 25, 2005 12:18 PM
© 2006 by Taylor & Francis Group, LLC

28

-1

28

The Electrolizing

Thin, Dense,

Chromium Process

28.1 General Definition

28-

1
28.2 Applications

28-

2

28.3

28.4 Solution

28-

5
28.5 Properties

28-

5

28.1 General Definition

The Electrolizing process uniformly deposits a dense, high chromium, nonmagnetic alloy on the surface
of the basic metal being treated. The alloy used in Electrolizing provides an unusual combination of
bearing properties: remarkable wear resistance, an extremely low coefficient of friction, smooth sliding
properties, excellent antiseizure characteristics, and beneficial corrosion resistance. Electrolized parts
perform better and last up to 10 times longer than untreated ones.
The solution and application processes are carefully monitored at all Electrolizing facilities. The result
is a fine-grained chromium coating that is very hard, thin, and dense and has absolute adhesive qualities.
The Electrolizing process deposits a 99% chromium coating on the basis metallic surfaces, whereas normal
conventional chromium plating processes tend to deposit 82 to 88% chromium in most applications.
Electrolizing calls for the cleaning and removal of the matrix on the basis metal’s surface by multi-
cleaning process, using a modified electrocoating process that causes the chromium metallic elements of
the solution to bond to the surface porosity of the basis metal. It is during this process that the absolute
adhesive characteristics and qualities of Electrolizing are generated. The Electrolizing coating will not
flake, chip, or peel off the basis metal substrate when conventional ASTM bend tests and impact tests
are performed. Three basic factors are always present after applying Electrolizing to metal surfaces:
•Increased wear (Rockwell surface hardness of 70 to 72 R

c

)
•Added lubricity characteristics
•Excellent corrosion resistance

Michael O’Mary

The Armoloy Corporation

DK4036_book.fm Page 1 Monday, April 25, 2005 12:18 PM
© 2006 by Taylor & Francis Group, LLC
General • Specific

Thickness • Adhesion • Corrosion • Wear Resistance (Surface
Hardness) • Lubricity • Conformity • Heat Resistance •
Surface Preparation 28-4
Brightness • Hydrogen Embrittlement

The Electrolizing Thin, Dense, Chromium Process

28

-3

•Automotive
•Business machines
•Cameras and projectors
•Computers
•Cryogenics
• Data processing
• Electronics
•Food processing
• Gauges and measuring equipment
•Medical instruments
•Metalworking
•Molds (plastic and rubber)
•Motor industry
•Nuclear energy
• Pharmaceutical
• Photography (motion and still)
•Refrigeration
•Textile industry
•Transportation

Specifically, Electrolizing is approved and meets the following aerospace, nuclear, and commercial
specifications:
•Air Research Company, Garrett, CO
•American Can Company
• AMS 2406
• AMS 2438
•AVCO Lycoming — AMS 2406
•Bell Helicopter
•Bendix Company
Utica, NY, division
Te terboro, NJ, division
Kansas City, MO, division
South Bend, IN, division
•Boeing
BAC 5709 Class II, Class IV
QQC 320
•Cleveland Pneumatic Tool-CPC Specs (Chromium), QQC320
•Colt Industries
Menasco, TX, division
•DuPont
•Fairchild Camera
•Fairchild Republic
•General Dynamics
•General Electric
Lynn, MA
Cincinnati, OH (aircraft)
Wilmington, MA
Wilmington, NC (nuclear)
Fitchburg, MA
• Gillette Company, Boston

•Grumman Aircraft

DK4036_book.fm Page 3 Monday, April 25, 2005 12:18 PM
© 2006 by Taylor & Francis Group, LLC

29

-1

29

The Armoloy

Chromium Process

29.1 General Definition

29-

1
29.2 Applications

29-

1

29.3 Surface Preparation

29-

2
29.4 Properties

29-

2

29.1 General Definition

The Armoloy process is a low temperature, multistate, chromium alloy process of electrocoating based
on a modified chromium plating technology. However, instead of the customary chromium plating
solutions, the Armoloy process uses a proprietary chemical solution. The solution and application process
are carefully monitored at all Armoloy facilities. The result is a satin finish chromium coating that is very
hard, thin, and dense and has absolute adhesive qualities. Armoloy deposits a 99% chromium coating o
the basis metallic surfaces, whereas conventional chromium plating processes tend to deposit 82 to 88%
chromium in most applications.
The Armoloy process involves cleaning and removing the matrix on the basis metal’s surface by special
proprietary means and using a modified electrocoating process that causes the chromium metallic
elements of the solution to permeate the surface porosity of the basis metal. It is during this process that
the absolute adhesive characteristics and qualities of Armoloy are generated. The Armoloy coating actually
becomes part of the basis metal itself, and the result is a lasting bond and a continuous, smooth, hard
surface. The surface will not chip, flake, crack, peel, or separate from the basis metal under conditions
of extreme heat or cold, or when standard ASTM bend tests are involved.
Three basic factors are always present after applying Armoloy to metal surfaces:
•Increased wear (70 to 72 R

c

surface hardness)
•Added lubricity characteristics (including the ability to utilize Armoloy against Armoloy)

•Excellent corrosion resistance

29.2 Applications

29.2.1 General Applications

All ferrous and most nonferrous materials are suitable for Armoloy application. Service life of parts has
been increased to 10 times normal life and even higher in certain applications. However, basis metals of
aluminum, magnesium, and titanium are not good candidates for the Armoloy process.

Michael O’Mary

The Armoloy Corporation

DK4036_book.fm Page 1 Monday, April 25, 2005 12:18 PM
© 2006 by Taylor & Francis Group, LLC
General Applications • Specific Applications
Thickness • Adhesion • Corrosion • Wear Resistance • Lubricity •
Embrittlement
Conformity • Heat Resistance • Brightness • Hydrogen

The Armoloy Chromium Process

29

-5

The plating cycle times are very short, and the Armoloy chrome is deposited so rapidly that Armoloy
seals the surface porosity of the basis metal before hydrogen ions can invade the surface of the basis
metal. However, if required, Armoloy can be and will be postplate heat treated to specification.

DK4036_book.fm Page 5 Monday, April 25, 2005 12:18 PM
© 2006 by Taylor & Francis Group, LLC

30

-1

30

Sputtered Thin

Film Coatings

30.1 History

30-

1
30.2 General Principles of Sputtering

30-

1
30.3 Sputter Deposition Sources

30-

3

30.4 Other Process Considerations

30-

8
30.5 Properties of Sputtered Thin Film Coatings

30-

8
30.6 Thin Film Materials

30-

9
30.7 Applications for Sputtered Thin Films

30-

9

30.8 Additional Resources

30-

10
Bibliography

30-

10

30.1 History

Sputtering was discovered in 1852 when Grove observed metal deposits at the cathodes of a cold cathode
glow discharge. Until 1908 it was generally believed that the deposits resulted from evaporation at hot
spots on the cathodes. However, between 1908 and 1960, experiments with obliquely incident ions and
sputtering of single crystals by ion beams tended to support a momentum transfer mechanism rather
than evaporation. Sputtering was used to coat mirrors as early as 1887, finding other applications such
as coating fabrics and phonograph wax masters in the 1920s and 1930s. The subsequent important process
improvements of radio frequency (rf) sputtering, allowing the direct deposition of insulators, and mag-
netron sputtering, which enables much higher deposition rates with less substrate damage, have evolved
more recently. These two developments have allowed sputtering to compete effectively with other physical
vapor deposition processes such as electron beam and thermal evaporation for the deposition of high
quality metal, alloy, and simple organic compound coatings, and to establish its position as one of the
more important thin film deposition techniques.

30.2 General Principles of Sputtering

Sputtering is a momentum transfer process. When a particle strikes a surface, the processes that follows
impact depend on the energy of the incident particle, the angle of incidence, the binding energy of surface
In sputtering, the incident particles are usually ions, because they can be accelerated by an applied
electrical potential. If the kinetic energy with which they strike the surface is less than about 5 eV, they

Brian E. Aufderheide

W. H. Brady Company

DK4036_book.fm Page 1 Monday, April 25, 2005 12:18 PM
© 2006 by Taylor & Francis Group, LLC

Direct Current Diode Sputtering • Triode Sputtering • Radio
Electrical • Magnetic • Optical • Mechanical • Chemical •
Decorative
Frequency Sputtering • Magnetron Sputtering • Beam
Sputtering • Reactive Sputtering
atoms, and the mass of the colliding particles (Figure 30.1).

Sputtered Thin Film Coatings

30

-5

“racetrack” effectively increases the number of ionizing collisions per electron in the plasma. The magnetic
confinement near the target results in higher achievable current densities at lower pressures (10

–3

to 10

–2

torr), nearly independent of voltage. This manner of cathode operation is described as the magnetron
mode and is capable of providing much higher deposition rates (10 times dc diode) with less electron
bombardment of the substrate and therefore less heating. Factors affecting deposition rate are power
density on the target, erosion area, distance to the substrate, target material, sputter yield, and gas pressure.
Dc is usually used for magnetron sputtering, but rf can be used for insulators or semiconductors. When
magnetic materials are sputtered, a thinner target is often necessary to maintain sufficient magnetic field
strength above the target surface. The three most common magnetron cathode designs, described below,
are illustrated in Figure 30.5.

30.3.4.1 Planar Magnetron

An array of permanent magnets is placed behind a flat, circular or rectangular target. The magnets are
arranged such that areas in which the magnetic field lines are parallel to the target surface form a closed
loop on the surface. Surrounding this loop, the magnetic field lines generally enter the target, perpen-
dicular to its surface. This produces an elongated electron racetrack and erosion pattern on the target
surface. Because of the nonuniformity in target erosion, utilization of target material is poor, typically
26 to 45%. This also results in nonuniform deposition on a stationary target. Uniformity is provided by
substrate motion, usually linear or planetary, combined with uniformity aperture shielding. Planar
magnetron cathodes are usually operated at 300 to 700 V providing a current density of 4 to 60mA/cm

2

or a power density of 1 to 36 W/cm

2

.

FIGURE 30.5

Clockwise from upper left: schematic representations of planar magnetron, gun-type magnetron, and
cylindrical post magnetron sputtering sources. (Adapted from J. A. Thornton, in

Deposition Technologies for Films
and Coatings

, R. F. Bunshah, Ed. Park Ridge NJ: Noyes Publications, 1982, pp. 194–195.)
?

?
?
Anode
Anode
Anode
Magnetic
Field
Magnetic
Field
Cathode
Cathode
Cathode
E × B
Electron
Motion
E × B
Electron
Motion
Plasma
Ring
Plasma
Ring
Plasma
Magnetic
Field Line
Primary
Electrons
Ultimate
Electrons

DK4036_book.fm Page 5 Monday, April 25, 2005 12:18 PM
© 2006 by Taylor & Francis Group, LLC

31

-1

31

Vapor Deposition

Coating Technologies

31.1 Introduction

31-

1
31.2 Physical Vapor Deposition

31-

3

31.3 Chemical Vapor Deposition

31-

16

31.4 Decorative and Barrier Coatings

31-

22

31.5 Conclusions

31-

28
References

31-

28

31.1 Introduction

For over 25 years, the thermal evaporation of aluminum onto thin polymeric webs, such as polyester
(PET) and polypropylene (PP), has generated large volumes of barrier packing films, decorative films,
capacitor films, and some window films. The experience of wide web handling was combined with
deposition technologies, such as electron beam evaporation, magnetron sputtering, and plasma-enhanced
chemical vapor deposition, to create a large number of new, exciting, coating materials, including oxides
and nitrides of most elements. More particularly, combinations of these coating layers into a complex
coating stack led to new products, such as low emissivity, solar heat reflecting, architectural glazing films,
electrochromic devices, and high-performance optical reflectors. With these technologies, unique coating
characteristics can be realized, e.g., transparent electrodes, flexible glassy barriers for moisture and gases,
and amorphous soft magnetic materials for security devices.
To build highly functional components with vacuum-coated webs, severe quality standards should be

in use today.

1

Many different characteristics are important for further functionality after coating, e.g.,
mechanical tensile strengths, Young’s modulus, surface finish, optical clarity (e.g., transparency haze),
and resistance to corrosion and UV irradiation. Many special additional surface treatments have also
materials that have been used industrially to date. Recently, the technical availability of low-cost SiO

2

and A1

2

O

3

coatings has created very interesting coating stack building blocks.

2

Since 1980, tool coatings formed by physical vapor deposition (PVD) technologies have become a
reality, and an industry has evolved around PVD tool coatings based on the work of the early pioneers
in this field.

3–6

Lindas Pranevicius

Vytautas Magnus University

DK4036_C031.fm Page 1 Thursday, May 12, 2005 9:40 AM
© 2006 by Taylor & Francis Group, LLC
Thermal Evaporation • Electron Beam Evaporation • Sputter
Arc Deposition
Thermal Chemical Vapor Deposition • Plasma-Enhanced
Deposition in Plasma • Reactive Sputter Deposition • Cathodic
Chemical Vapor Deposition (PECVD)
Decorative Coatings • Barrier Coatings
employed for the flexible substrates used for vacuum coating. Table 31.1 gives a survey of most substrates
become indispensable in attaining necessary product performance. Table 31.2 lists most of the coating

Vapor Deposition Coating Technologies

31

-5

sibility of deposition at relatively low substrate temperatures. The major roles of the plasma in various
plasma-assisted processes are related to activation and enhancement of the reactions that are necessary
for deposition compound films, and modification of the growth kinetics and, hence, modification of the
structure and morphology of the deposits.
Due to the above consideration, plasma is used in a variety of physical and chemical vapor deposition
processes.
The most commonly used techniques for plasma-assisted PVD are as follows: (1) sputtering, including
direct current (dc), radio frequency (rf), triode, or magnetron geometries and reactive sputtering using
dc, rf, triode, or magnetron sources; and (2) activated reactive evaporation.
The presence of the plasma in the source–substrate space significantly affects the processes occurring

at each of these steps in film deposition, which are generation of species, transport from source to
substrate, and film growth on the substrate.
Moreover, the effect of the plasma on the above three steps differs significantly between various
processes. Such differences are manifest in terms of the types and concentrations of the metastable species,
ionized species, and energetic neutrals that, in turn, influence the reaction paths or steps involved in the
overall reaction of film formation and the physical location of these reaction sites. Moreover, it should
be noted that the ionizing probability is maximum for electrons in the range of 50 to 60 eV and decreases
with further increase in energy. It is, therefore, advantageous to have low-energy electrons for ionization
of the gas and vapor species.

31.2.3.1 Diode Plasmas

The dc-diode plasma device is the simplest form of a plasma used for sputtering and sputter deposition.
The system consists of a cathode, an anode, a dc power supply, and an enclosure. The interrelation
between gas density, electrode spacing, and applied voltage needed for the breakdown of the gas and the
formation of plasma is given by Paschen curves.

17

Only a tiny fraction (about 0.01%) of the gas atoms
are ionized — the majority are neutral. The electrons in the plasma are relatively hot, with a Maxwellian
energy distribution and an equivalent thermal temperature of 10,000 to 50,000 K. The electron temper-
ature is usually described with energy units (eV), where 1 eV is about 11,600 K.
Because the plasma is conductive, there is virtually no potential gradient with the plasma itself. All of
the electric fields occur at the edge of the plasma in a region called a sheath. Due to the large proportion
of neutral gas atoms to ions, the ions are in thermal equilibrium with the gas atoms (through collisions)
and are only at a temperature in the range of 100 to 1000

°

C. And due to the much higher electron
temperature and lower mass, the electrons move rapidly around the plasma. This last effect results in
the appearance of several different potentials within the system.
The plasma potential is the apparent voltage on the bulk of the plasma away from the sheath. The
floating potential is the potential reached by an electrically isolated object immersed in the plasma. It is
also the potential (on any surface, conductive or not) at which the arriving ion and electron fluxes are
equal. The floating potential is always the negative of the plasma potential, typically by a factor 3 times
the electron temperature.
For objects floating electrically in the plasma, the energy is usually less than 20 eV and causes little
sputtering. For a surface such as the cathode, the ion energy is equal to the difference between the plasma
potential (a few more volts positive than the anode) and the cathode voltage. These energies can be
several hundreds of eV and will cause significant sputtering of the cathode surface. Therefore, a sample
to be coated with a thin film of sputtered atoms could be located on the anode surface or virtually
anywhere within the chamber.
Dc-diode plasmas are characterized by low etching and deposition rates. The reason for the low rates
is a low plasma density due to a cross for electron-impact ionization that is fairly small. Therefore, to
get a high plasma density and, hence, a high ion bombardment rate, the gas pressure must be increased
to pressures near 133 Pa. In addition, the voltages needed for moderate currents are fairly high, several
kV. The resultant sputtered atoms are rapidly scattered by the background gas, and the net deposition
rate on a sample surface is fairly low.

DK4036_C031.fm Page 5 Thursday, May 12, 2005 9:40 AM
© 2006 by Taylor & Francis Group, LLC

31

-6

Coatings Technology Handbook, Third Edition

Dc-diode sputtering is also constrained by the requirement that the electrodes must be metallic
conductors. If one of the electrodes is insulating, it charges rapidly, and additional current is suppressed.
This effect can occur if a reactive gas, such as oxygen or nitrogen, is introduced into the plasma, resulting
in the oxidations of metal surfaces on the electrodes. Therefore, dc-diode sputtering is not an appropriate
technology for the deposition of most compounds and dielectrics.
By operating the plasma diode with an ac potential, rather than dc, these problems can be overcome
(Figure 31.1). At the most commonly used frequency of 13.6 MHz, there is little voltage drop across the
insulating electrode or layer. The electrodes will not charge up, and therefore, it is possible to sputter
dielectrics or reactively sputter metals. There is an additional degree of ionization with an rf-powered
plasma due to additional energy transmitted to the plasma electrons at the oscillating sheath. The net
result is a higher plasma density, compared to dc-powered plasmas, and the ability to operate at lower
system pressures (0.5 to 120 mPa).
The cathode of a typical rf-diode system is usually powered through an impedance-matching device
known as matchbox. The function of the matchbox is to maximize the power flow from the rf generator,
which has an output impedance of 50 ohms, to the plasma, which has a complex impedance usually in
the 1000 ohm range. A series capacitor is included in the matchbox to allow the formation of a dc bias
on the cathode. This occurs due to the higher electron mobility and results in a negative dc potential
on the powered electrode of up to one-half applied rf peak-to-peak voltage. The ions in the plasma that
are accelerated to the cathode are too massive to respond to the 13.6 MHz fields and respond only to
the dc bias.
The most common application of rf-diode sputtering is for the deposition of dielectric films. Often,
the sample surface is biased slightly during the deposition to provide some level of ion bombardment
that results in changes to the density and microstructure of the films and some degree of resputtering
that leads to increased planarization.

31.2.3.2 Magnetically Enhanced Plasmas

Electrons in a magnetic field are subjected to Lawrence force, which in a homogeneous magnetic field
perpendicular to the electron motion would cause the electron to move in a circular path with radius,
known as the Larmor radius. In the direction of the magnetic field, there is no net magnetic force, so

the electrons are unconfined. The net result is that electrons tend to spiral along magnetic field liners in
a helical path. By constraining the electron to this motion, the effective path length of the electron is
increased significantly, and hence, the probability of an ionizing electron–atom collision is increased. For

FIGURE 31.1

The rf excitation system:

R

a

and

R

c

— anode and cathode sheath resistances;

R

p

— plasma resistance;

C

a

and

C

c

— the geometric sheath capacitances.
Plasma
Cathode Sheath
Anode Sheath
Cathode
Anode
R
p
R
a
C
a
D
a
R
c
C
c
D
c
rf
rf

DK4036_C031.fm Page 6 Thursday, May 12, 2005 9:40 AM

© 2006 by Taylor & Francis Group, LLC

Vapor Deposition Coating Technologies

31

-7

a given applied power, then, the effect of the magnetic field is to reduce the plasma impedance, resulting
in higher discharge currents at lower voltages. The increased density also allows significant reductions
in the background pressure, such that the magnetically enhanced plasmas can operate at pressures as low
as the 10

–2

Pa range.
Magnetrons are the most common form of magnetically enhanced plasmas. In the device, a magnetic
field is configured to be parallel to the surface of the cathode. There is a resulting electron drift, caused
by the cross-product of the electric and magnetic fields (known as an E-cross–B-drift) that tends to trap
electrons close to the cathode surface. The drift motion is directional, and in a magnetron it is configured
to close on itself. A common example of this is shown in Figure 31.2, for a circular geometry, and is
called a circular planar magnetron. In this case, the magnetic field is configured to be radial pole and a
perimeter, or ring magnetic pole.
The magnetron device, which is defined as having a closed-loop E

×

B drift path for the secondary
electrons, has been developed in a number of geometries.

18

Perhaps the most common alternative is to
use a rectangular configuration, known as a “racetrack” magnetron (Figure 31.3). This geometry has
some intrinsic advantages for the automated handling of parts.

FIGURE 31.2

Circular planar magnetron showing an expanded view of the pole-piece configuration. Water cooling
is not shown, but it typically occupies the volume between the cathode and the back of the pole-piece assembly.

FIGURE 31.3

A rectangular or racetrack magnetron.
S
N
Cathode
Magnetic Field
Lines
E × B Drift Path
N
S
Magnetic Pole Piece Assembly
Cathode
E × B Drift Path

DK4036_C031.fm Page 7 Thursday, May 12, 2005 9:40 AM
© 2006 by Taylor & Francis Group, LLC

31

-8

Coatings Technology Handbook, Third Edition

Magnetron plasmas have a unique feature in that the secondary electrons are strongly constrained
to the region near the cathode surface. This causes dense plasma to form near the cathode in the region
of the drift loop. The dense plasma results in very high levels of ion bombardment of the cathode surface
and, hence, high rates of sputtering. The high-rate ion bombardment is localized on the cathode directly
under the E

×

B path. The resultant sputter emission of atoms is also localized, which means that
deposition uniformity is usually not good. Therefore, for most deposition systems, it will be necessary
to either move the sample or alter the magnetron location to attain good deposition uniformity. In
addition, the erosion of the cathode is also localized, which results in poor utilization of the cathode
material, as deep grooves are eroded into the cathode surface in the vicinity of the E

×

B drift path.
The wide grooves are called the “etch track” and are characteristic of magnetron sputtering. Typically,
only 10 to 15% of the cathode material can be used before the grooves start to etch through the back
of the cathode.

19

At high pressures, the distribution of sputtered atoms is smeared out due to gas scattering and
deposition homogeneity increases, but the cost is a real reduction in the sputtered atom’s kinetic energy

and a potential change in the film properties. There are two obvious solutions to the nonuniformity. The
first is to move the sample in some way to average the deposition over the sample surface. For circular
planar magnetrons, this requires a fairly complicated planetary motion. The alternative, using conven-
tional rotating-sample motion, is to use deposition shields located between the cathode and the sample,
which effectively collect the sputtered flux locally. This process, however, reduces the net deposition rate
over the entire sample to the lowest level of the original distribution.
For rectangular or other elongated magnetrons, the solution for increased uniformity is to translate
the samples past the magnetron perpendicular to the “long” direction of the cathode. An example is
shown in Figure 31.4, which shows a rectangular magnetron system viewed end-on, in which the samples
move from one end of the system to the other. The dimensions of these systems on a manufacturing
scale can be rather large. A common size uses cathodes 2 m in length in a sputtering system with an
overall length that exceeds 20 m.

20

For some industrial applications, in particular those where contamination is a critical concern, it may
not be desirable to move the samples during deposition. In this case, magnetrons have been developed
that have a moving etch track.

21

Over time, the eroded area is fairly uniform, and a high degree of
uniformity can be obtained when depositing films on large, stationary substrates. The moving etch track
is set up by rotating the magnet assembly in the cooling water behind the cathode face. An industrial
cathode of this design might have a diameter of 25 cm and be rated at a power of 25 kW. The second
important advantage of these magnetrons is that the utilization of the cathode is very efficient: up to
80% of the cathode material can be used for sputtering, compared to 15% for a nonrotating magnetron.
This results in much better efficiency and longer time periods between cathode changes. Because of this

FIGURE 31.4

An automated in-line system based on rectangular planar magnetrons.
Samples
To High
Vacuum Pump
To Load-lock
Pump
To Load-lock
Pump
Load Un-load
Valve Valve
Magnetrons

DK4036_C031.fm Page 8 Thursday, May 12, 2005 9:40 AM
© 2006 by Taylor & Francis Group, LLC