Professional Documents
Culture Documents
Electrophoretic Deposition of Materials
Electrophoretic Deposition of Materials
Electrophoretic Deposition of Materials
ELECTROPHORETIC DEPOSITION
OF MATERIALS
Omer O. Van der Biest and Luc J. Vandeperre
Departement Metaalkunde en Toegepaste Materiaalkunde, Katholieke Universiteit
Leuven, De Croylaan 2, 3001 Heverlee, Belgium;
e-mail: Omer.Vanderbiest@mtm.kuleuven.ac.be
ABSTRACT
The electrophoretic deposition of materials is reviewed. Numerous applications
of electrophoretic deposition are described, including production of coatings,
free-standing objects, and laminated or graded materials, infiltration of porous
materials, and fabrication of woven fiber preforms. The preparation of elec-
trophoretic suspensions is discussed as are a number of mechanisms of deposi-
tion that have been proposed elsewhere. In discussing the kinetics of the process,
primary attention is given to the relation between the evolution of the current and
the electric field strength.
INTRODUCTION
Electrophoretic deposition (EPD) is essentially a two-step process. In the first
step, particles suspended in a liquid are forced to move toward an electrode
by applying an electric field to the suspension (electrophoresis). In the second
step, the particles collect at one of the electrode and form a coherent deposit
on it. It should be noted that the process yields only a powder compact, and
therefore electrophoretic deposition should be followed by a densification step
such as sintering or curing in order to obtain a fully dense material.
The particles in suspension will move only in response to the electric field if
they carry a charge. Four mechanisms have been identified by which the charge
on the particles can develop (1): (a) selective adsorption of ions onto the solid
particle from the liquid, (b) dissociation of ions from the solid phase into the
liquid, (c) adsorption or orientation of dipolar molecules at the particle surface,
327
0084-6600/99/0801-0327$08.00
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
and (d) electron transfer between the solid and liquid phase due to differences
in work function.
A charged particle in a suspension is surrounded by ions with an opposite
charge in a concentration higher than the bulk concentration of these ions; this
is the so-called double-layer (Figure 1). When an electric field is applied, these
ions and the particle should move in opposite directions. However, the ions are
also attracted by the particle, and as a result, a fraction of the ions surrounding
the particle will not move in the opposite direction but move along with the
Figure 1 Schematic of the double layer surrounding a charged particle and evolution of the electric
potential from the surface potential, ψ0 , to zero far from the particle. The potential at the surface
of shear, the limit between the liquid moving with the particle and the liquid, which does not
move with the particle, is termed the zeta-potential, ψζ , and is the main parameter determining the
electrokinetic behaviour of the particle.
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
particle. Hence, the speed of a particle is not determined by the surface charge
but by the net charge enclosed in the liquid sphere, which moves along with
the particle. The potential at the surface of shear is termed the zeta-potential or
electrokinetic potential. In principle, a particle with a negative surface charge
can show a positive zeta-potential. The latter occurs, for example, when the
charge of specifically adsorbed ions is higher than the surface charge.
In fact, the equilibrium speed of the particle is determined by four forces act-
ing on the particle. The first, which accelerates the particle, is the force caused
by the interaction of the surface charge with the electric field. All other forces
slow the particle: These are viscous drag from the liquid following Stoke’s law,
the force exerted by the electric field on the counter-ions in the double layer (re-
tardation) and, when a particle moves, the distortion in the double layer caused
by a displacement between the center of the negative and positive charge, (re-
laxation). A more complete treatment of these forces, along with information
on calculations of their magnitudes, can be found, for example, in Reference 2.
In this paper, we first present a number of applications of electrophoretic
deposition. Next, the preparation of suspensions for electrophoretic deposi-
tion is discussed, followed by a synthesis of the available information on the
mechanisms of electrophoretic deposition and the kinetics of the process.
APPLICATIONS
Electrophoretic deposition can be applied to any solid that is available in the
form of a fine powder (< 30 µm) or a colloidal suspension. Examples of elec-
trophoretic deposition of materials of almost any material class can be found,
including metals, polymers, carbides, oxides, nitrides, and glasses (see Table 1).
Furthermore, the process can be used for producing coatings, for shaping mono-
lithic, laminated and graded free-standing objects, and for infiltration of porous
materials and woven fiber preforms for composite production.
Coatings
Electrophoretic deposition of coatings has already gained a world-wide accep-
tance for automotive, appliance, and general industrial (organic) coatings (3).
For example, the construction and operation of an 800 bodies-per-hour cat-
aphoretic coating plant by a German car builder was described recently (4).
The high level of automation, low levels of pollution, high throw power and
homogeneity of the coating are the advantages that have led to the use of elec-
trophoretic deposition (3). Other advantages are the improved adherence of
electrophoretically deposited coatings and their higher density compared with
dipped or sprayed coatings (5).
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
Table 1 A non-exhaustive list of examples from a broad range of material classes that have been
electrophoretically depositeda
Acids and hydroxides Calcium hydroxide (91), magnesium hydroxide (91), antimonic acid
(96), boric acid (52)
Borides LaB6 (97)
Carbides B4C (36), C (34–36, 98), Cr3C2 (99), NpC (88), diamond (100, 101),
PuC (88), ThC (88), UC (88), SiC (34–36, 57, 98, 102), UWC2 (88),
WC (97)
Carbonates (87, 91, 97, 103)
Metals Al (83, 97, 104), Al-Cr (65), Al-Si (65), Al-Ti (65), Au (88, 105),
B (88), Cu (83), Dy (88), Fe (83), Mo (88, 97), MoSi2 (51), Nb (88),
Nb2Sn (88), Ni (83, 97), Sn (83), Re (88), Ru (97), W (88, 97),
Zn (83, 97), Zr (88)
Nitrides Si3N4 (56), AlN (37)
Organic materials Starch (47), styrene-acrylic copolymer (106), latex (3, 85), vinyl
copolymers (3), epoxy resins (3), polyamides (3), poly-urethanes (3)
Oxides Clay (25, 26), Al2O3 (30, 31, 48, 87, 97, 98), ß-Al2O3 (39, 64), BaTiO3
(107), Cr2O3 (108), Fe2O3 (108), glass (17, 109), In2O3 (97), La2O3
(51), LiAlO2 (110), mica (111), MgO (97, 112), NiO (88, 108),
ReBaCuO (24), SiO2 (53, 58, 59, 97), TiO2 (97, 108), UO2 (88),
YBaCuO (18–20), ZnO (97), ZrO2 (41, 44, 48, 51, 54, 85)
Phosphors (14–16, 81, 97, 113, 114)
a
Numbers in parentheses are references
showing that although EPD can be a fast process, good control of the growth of
the layers can be obtained. The same group also produced alumina/lanthanum
aluminate laminates (42). Ferrari et al (43) and Fischer et al (44) produced alu-
mina/zirconia laminates from aqueous suspensions. Vandeperre & Van der Biest
made a range of SiC-based laminates with graphite (45), SiC+graphite (46),
and porous SiC interlayers (47). They also produced laminated SiC/graphite
composite tubes (35), thus combining the ability of EPD to produce laminated
materials with the shaping capabilities of the process.
GRADED MATERIALS While layered materials are obtained by immersing the
deposition electrode in different baths, graded materials can also be made by
gradually changing the composition of the suspension from which EPD is car-
ried out. Sarkar et al demonstrated the ability to form graded materials by slowly
adding an ethanol-based suspension of an alumina powder to an ethanol-based
suspension of an yttria-stabilized zirconia powder during deposition (48). A
gradual increase in the alumina content of the deposit was observed. Later,
the same group also produced Al2O3/MoSi2- and Al2O3/Ni-graded materials
(49). Chao et al produced alumina/ceria-stabilized zirconia-graded rods from
an acetone-based suspension (50).
UNIQUE MICROSTRUCTURES AND SELECTED DEPOSITION The full potential of
EPD in producing unique microstructural features is perhaps not yet realized.
Nicholson et al (51) did some exploratory work in making nonplanar laminates:
By placing a grid before the deposition electrode, they were able to produce
laminates with wavy interlayers. Thus by using auxiliary electrodes, deposition
can be enhanced locally. Scala & Sandor (52) have deposited B2O3 on silicon
wafers as a boron diffusion source. The deposition onto the entire silicon wafer
was prevented by growing a silica film on the wafer and etching the silica away
only where deposition of boron was required.
Infiltration
The throwing power of electrophoretic deposition can also be used to infiltrate
objects with a matrix material or to apply an internal coating. Gal-Or et al (53)
infiltrated porous graphite electrodes with silicon carbide and silicon oxide
particles. For a good infiltration, the particle buildup in an external coating
should be decreased so that the pores are not blocked. Lower applied voltages
and a high-particle concentration in the suspension improved the infiltration.
Ishihara et al (54) applied an internal coating of yttria-stabilized zirconia to
a porous Ni-CaO stabilized ZrO2 cermet containing 40 wt% Ni for a solid
oxide fuel cell. By repetitive infiltration and sintering, a dense film of uniform
thickness could be applied.
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
when, for example, multivalent metal ions, which adsorb specifically on the
surface, are added to the suspension (61).
Because of the lower solids loading, simply changing the pH to a value far
from the pzc using concentrated acids or bases is generally all that is required
for obtaining a stable suspension in water. Naturally, the ionic concentration
must be kept low as the stability decreases with increasing ionic concentration
(Figure 2), and the pH range must be adapted so that the oxide does not dissolve.
For example, yttria dissolves readily in an acid aqueous environment.
However, the use of water-based suspensions causes a number of problems in
electrophoretic forming (1). Electrolysis of water occurs at low voltages (∼5 V),
and therefore gas evolution at the electrodes is inevitable at field strengths high
enough to give reasonably short deposition times. If electrophoresis is used to
form objects, the inclusion of gas bubbles in the deposit can be prevented by
depositing on a porous membrane placed before the electrode. However, current
densities are higher compared with non-aqueous media, which leads to Joule
heating of the suspension and sometimes loss of stability of the suspension.
Figure 2 Interaction energy (attraction is negative) between two particles surrounded by a dou-
ble layer for a surface potential of 25 mV, a particle radius of 0.35 µm, a Hamaker constant of
10−20 J, and a relative dielectric constant of 20, for various concentrations of a background elec-
trolyte (indicated in the figure). With increasing electrolyte concentration, the height of the repul-
sion barrier due to double layer overlap decreases, and the repulsion vanishes completely at even
higher electrolyte concentration.
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
between powder dispersability and the difference of the iep of the powder in
water and the pKa or the pKb value of the solvent.
Given the importance of the pH in determining the charging of powders in
water, it was only logical that parameters describing the acidity/alkalinity of
non-aqueous media were used in the first attempts to understand charging in
non-aqueous media. However, one important difference has not attracted much
attention: The charge on a powder in aqueous suspensions is not determined
by the pH of the water before the suspension is prepared, but by the pH of the
suspension once prepared. This simple fact is not stressed when dealing with
aqueous suspensions because the pH is regarded as a tuneable experimental
parameter. Indeed, only limited amounts of concentrated acids or bases are
required in practice to tune the pH. Hence, changes in pH upon addition of a
powder to water are given almost no attention. That the pH can change substan-
tially and unpredictably when a powder is added to water becomes apparent if
the natural pH values of several powders are compared. The term natural pH
of a powder is used in this review to describe the pH at which the suspension
equilibrates when a powder is suspended in water. For example, in two batches
of a nominally identical SiC powder, one had a natural pH of 3.2 at 20 g · L−1,
whereas the other had a natural pH of 8.2 at 20 g · L−1. A more systematic in-
vestigation showed that simply washing powders in an acid or alkaline aqueous
suspension for a short time suffices to adsorb acids or bases onto the powder,
which results in a remarkable difference in charging and deposition behavior in
acetone, ethanol, and iso-propyl alcohol (72, 73). As a result, the charge on the
powder when added to a suspension medium will not only be correlated with
the difference in pzc and the acidity of the medium, but also with the amount of
acids and bases adsorbed on the powder as expressed by the natural pH. Even
for a clean powder surface, the natural pH of a suspension tends toward the
pzc of the powder (72). Figure 3 illustrates these two effects for a number of
commercially available powders (74).
If it is likely that ions will leach from the solid to be deposited, as e.g. Na+
ions from beta-alumina (75), or barium ions from barium titanate powders (76),
the design of the suspension must take into account that the ionic conductivity
will increase, that the solid might be charged through the dissolution of these
ions, and that re-adsorption of these ions can also influence the zeta-potential.
MECHANISM OF DEPOSITION
Although electrophoretic deposition is an old process, the exact mechanisms
that allow a deposit to be formed are still not entirely clear. According to
Hamaker & Verwey (77), the formation of a deposit by electrophoresis is akin to
the formation of a sediment due to gravitation. The pressure exerted by incoming
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
layer of a particle in motion is distorted so that the double layer is thinner ahead
of the particle. Further, they argue that since counter ions move along with the
particle, their concentration in the deposit is much higher than normally would
be in the suspension. Thus they probably recombine with incoming co-ions to
reform their original salts, which results in a thinning of the double layers of the
particles in the deposit. Thus an incoming particle with a thinner double layer
meets with the particles in the deposit that also have a thinned double layer
and as a result the repulsion is decreased substantially, allowing the particles to
overcome this repulsion barrier.
A type of suspension probably best documented with respect to the deposi-
tion mechanism is one in which the powder is charged positively by adsorption
of metal ions on the powder. These metal ions are introduced in the suspension
through addition of salts such as Mg(NO3)2, La(NO3)3, Y(NO3)3, MgCl2, or
AlCl3. If a current is passed through solutions of these salts, the formation of
a hydroxide (16, 81, 82) is observed. If the water content of such solutions in
iso-propyl alcohol is limited to below approximately 5 vol%, the main elec-
trochemical reaction product becomes an alkoxide for Mg(NO3)2 or Al(NO3)3
(81). Analysis of deposits made from such suspensions has revealed that these
hydroxides also form during deposition (14). Therefore, it is likely that bind-
ing by these hydroxides is the mechanism of deposition. Hydroxides provide a
better adhesion than alkoxides (81), and a limited amount of water is thus an ad-
vantage for obtaining well-adhering coatings. Consequently, the deposition of
the powder is induced by an electrochemical reaction at the electrode, and one
would expect that no deposition would occur away from an electrode. Indeed,
Brown & Salt (83) had already observed that an aluminium powder charged
through addition of AlCl3 to ethanol did not deposit on a porous membrane
placed away from the electrode, whereas a deposit could be obtained from this
suspension on the cathode.
However, in the absence of such salts, electrophoretic deposition can oc-
cur away from the electrode. For example, magnesium oxide suspended in
methanol was found to deposit on a porous membrane placed before the elec-
trode (83), and aluminium oxide particles suspended in ethanol with addition
of an acid deposited on a dialysis membrane placed before a graphite cathode
(84). Thus although the mechanism of hydroxide formation is supported with
a substantial amount of experimental evidence, it clearly is not the only mecha-
nism for formation of a deposit. When Koelmans (78) proposed his mechanism
of flocculation (discussed above), he had already noted that electrolysis could
lead to insoluble or non-ionized products and therefore not to the increase in
the electrolyte concentration near the electrode required for his mechanism.
However, he also argued that for oxides, protons and hydroxyl ions are po-
tential determining ions. Therefore, if during electrolysis their concentration
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
changes near the electrode, the surface charge and the interparticle repulsion
will decrease. Thus the deposition could also be induced by a shift in pH at
the electrode. The fact that, in the absence of salts such as AlCl3, deposits can
be formed on membranes placed away from the electrode is not necessarily a
contradiction for such a mechanism. Sheppard (85) pointed out in his discus-
sion of electrophoretic deposition of rubber on membranes that generally in
such experiments the side of the membrane toward the anode becomes more
alkaline, whereas the side toward the cathode becomes more acidic. The con-
comitant increase in hydrogen ions is, in his opinion, a considerable factor in
the deposition of rubbers on such membranes. Note that if deposition is caused
by a change in pH, a critical time will be required before the pH has changed
sufficiently. Pierce (3) discussed the cathodic deposition of resins quite simi-
larly. He postulates that in order for a resin with amine groups to deposit, the
hydroxyl ion concentration at the electrode should reach a critical value related
to the solubility product for the ammonium hydroxide:
£ ¤
RESIN − NH+ −
3 [OH ] = K sp . 2.
Further, in the absence of migration and chemical reactions, the concentration
of hydroxyl ions at the cathode will increase as the square root of time for an
unstirred electrolyte at constant current density. Therefore, the product of the
square root of the critical time and the current density is expected to be constant
for such systems, which is also found experimentally (3).
To summarize, there are many indications that in order for powder to be
incorporated in the deposit, the charge on the surface must be reduced. For
powders charged by adsorption or de-sorption of protons, a different pH in
the suspension, and near the deposition electrode or the membrane on which
deposition is performed, could reduce the surface charge. For powders charged
by adsorption of metal ions, an electrode reaction converting these adsorbed ions
into uncharged metal hydroxides is required. Hence, no deposition is observed
on membranes in such systems. These hydroxides improve the coherence of
the deposit and the adherence to the electrode. Therefore using salts to charge
powders is probably better for making coatings than for making free-standing
objects, which have to be released from the electrode. If other ways are used to
charge a powder, for example by adsorption of a poly-electrolyte (33), whether
deposition has occurred will be determined by the decharging reactions required
for the surface groups of these poly-electrolytes.
Although the decharging due to changes in pH as a consequence of the
current is very appealing, it should be noted that such changes normally do
not extend very far into the liquid. However, it is possible that the change in
the contribution of convection to the transport of ions to the electrode through
the growth of the deposit causes the changes in pH to extend farther into the
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
KINETICS OF DEPOSITION
Single-Powder Deposition
The kinetics of electrophoretic deposition has been the subject of numer-
ous investigations. By simply applying the principle of conservation of mass,
Hamaker (77) proposed the following equation
dY
= f.µ.E.S.c. 3.
dt
with Y, yield of deposition (kg); t, deposition time (s); µ, electrophoretic
mobility (m2 · V−1 · s−1); S, surface area of the electrode (m2); c, solids con-
centration (kg · m−3); and f a factor taking into account that not all powder
brought to the electrode by electrophoresis is also incorporated in the deposit
( f = 1 or less). This equation is now accepted as describing the basis for the
kinetics of electrophoretic deposition.
The surface area of the electrodes and the concentration of solids in a sus-
pension can easily be determined or calculated during the process. The elec-
trophoretic mobility can also be measured. Although a good qualitative notion
of the electric field strength can be obtained through measurement (86), the
electric field strength is almost always evaluated from the applied voltage and
the inter-electrode distance. However, such a calculation is subject to condi-
tions. The potential drop over an electrophoretic deposition cell with flat equal
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
Figure 4 Evolution of the potential in an electrophoretic deposition cell: The applied potential is
consumed by a potential drop at each electrode and by an ohmic potential drop over the suspension
and the deposit.
where Va is the applied potential (V); I is the current (A); Rd and Rs are the
resistance per unit length of the deposit and of the suspension, respectively
(Ohm/m); d is the electrode separation (m); d1 the thickness of the deposit
(m); and 1φi is the potential drop at one electrode (V). The resistance per
unit length is defined as the specific resistance (Ohm/m) divided by the surface
area of the electrodes (m2). The conditions required in order for the electric
field strength in a suspension to be close to the applied potential divided by the
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
inter-electrode distance are thus: The potential drops at each of the electrodes
should be negligible, and the specific resistance of the deposit should be of the
same order as the specific resistance of the suspension.
The validity of the approximation for the electric field strength is rendered
even more doubtful by the fact that at constant applied potential, the current
is frequently found to decrease during deposition. The following explanations
have been proposed: a change in polarization at the electrode (87, 88), a higher
specific resistance of the deposit compared with the suspension (17, 30, 57,
89–92), and changes in the conductivity of the suspension (72, 93).
Starting at the electrode, one should realize that in order for a net current to
flow through the solid/liquid interface, a potential in excess of the equilibrium
potential is required (activation overvoltage). The relation between the excess
potential over the interface and the resulting net current is exponential (94).
Therefore, small variations in activation overvoltage (order of 1 V) result in
large changes in current. Hence, the activation overvoltage can be expected
to remain more or less constant during the process when depositing from non-
aqueous suspensions, where the applied voltage is typically of the order of
100–1000 V. A second effect to be taken into account is that according to
Nernst’s law, the equilibrium potential over a solid/liquid interface depends on
the concentration of the reactants (94):
µ ¶
RT aox
1φ = 1φ0 + ln = 1φ0 + ηc , 5.
zF ared
where 1φ0 is the standard equilibrium potential drop over the solid/liquid in-
terface (V), 1φ is the equilibrium potential drop over the solid/liquid interface
(V), aox is the activity of the species being oxidized, ared , is the activity of
the species being reduced, R is the gas constant (J · mol−1 · K−1), T is the
temperature (K), z is the valence of the reduction reaction, and F is Faraday’s
constant (C · mol−1). Changes in polarization due to a change in concentration
of the reactants for the electrode reaction are often referred to as concentration
overvoltage (ηc ). If during electrophoretic deposition the transport of reactants
for the electrode reaction toward the electrode is slower than the consumption
at the electrode, the concentration at the electrode will drop and the concen-
tration overvoltage will increase. Note that changes in the proton or hydroxil
concentration were expected and observed in some of the investigations into
the mechanism of deposition (3). Moreover, the fact that a deposit forms on
one of the electrodes can be expected to hinder the transport of reactants toward
that electrode, thus enhancing the probability for an increase in concentration
overvoltage (see Figure 5). Cathodic deposits are especially prone to serious
changes in concentration overvoltage. Species capable of being reduced must
travel from the solution toward the cathode. At the anode, one of the electrode
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
Figure 5 Illustration of a possible influence of the growing deposit on polarization at the electrode.
The concentration of the reactant for the electrode reaction C as a function of the distance X from
the electrode is shown for four conditions: (a) Before any potential is applied, no concentration
gradients exist. (b) As soon as a current starts to flow, the concentration of the species reacting at the
electrode drops by a small amount, and the amount consumed at the electrode is replaced through
diffusion, convection and migration. (c) The presence of the deposit influences the transport of
reactants to the electrode. If this results in a decreased incoming flux, the concentration at the
electrode will start to drop. (d) Because the diffusion flux is determined by the slope of the
concentration gradient, the diffusion term in the flux decreases with increasing deposit thickness
reactions can be the dissolution of the electrode material. Naturally, in that case
the reactant concentration for the oxidation reaction will not drop. Some types
of electrodes, however, can also suffer from a serious increase of the poten-
tial drop at the electrode, even when they are used as an anode. For example,
aluminum, when polarized anodically, does not dissolve but forms a protective
oxide layer. Not only does this layer introduce an extra electrical resistance,
the oxidation of new aluminium becomes increasingly more difficult as oxygen
has to travel through the oxide layer. Hence, other oxidation reactions can be
required, and these can require the transport of species from the solution to the
anode. Note that even when the deposit forms on the anode, an increase in
cathodic concentration overvoltage can also occur. For example, during anodic
deposition of SiC from acetone, Vandeperre & Van der Biest (72) found that the
potential drop at the cathode increased during deposition, resulting in a drastic
increase of the total resistance of the deposition cell (Figure 6).
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
Figure 6 (a) Evolution of the apparent resistance (applied potential divided by the measured
current, Rapp) during deposition of SiC from acetone and of the resistance of the suspension, Rsus,
as determined from a measured difference in potential drop between two reference electrodes
immersed in the cell and the measured current. Initially the magnitude of the current is controlled
by the resistance of the suspension, but after a few minutes, concentration polarization at the
counter electrode becomes very important, and the linear relation between amount of powder that
has deposited and the apparent resistance is lost. Initially the magnitude of the current is controlled
by the resistance of the suspension, but after a few minutes, concentration polarization at the
counter electrode becomes very important, and the linear relation between amount of powder that
has deposited and the apparent resistance is lost. (b) Evolution of the potential drop over the
suspension (Vsusp), at the counter electrode (Vcount) and at the deposition electrode (Vdep) for the
same deposition experiment
In regard to the conductivity of the suspension and the deposit, many authors
claim that the deposit has a much higher specific electrical resistance than does
the suspension. In this review the term resistance of the deposit is used only
to indicate the ohmic potential drop over the deposit as defined in Equation 4
and does not include possible effects of deposit formation on the concentration
overvoltage discussed above.
Most authors have simply inferred the higher specific resistance of the deposit
from the decrease in current (17, 30, 57, 89–92). Sarkar et al (86) measured the
electric field strength by measuring the potential drop between two platinum
probes placed between the electrodes and also monitored the potential drop
between the deposition electrode and one platinum probe in the suspension.
During cathodic deposition of alumina from ethanol at constant applied current,
these authors found that the electric field strength remained constant and that the
potential drop between the deposition electrode and the probe in the suspension
increased at the same rate as the voltage required to keep the current constant
increased. Since both the required increase of applied voltage, as well as the
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
growth of the deposit, were linear with time, they concluded that this result
was proof for the higher specific resistance of the deposit. However, changes
in the concentration overvoltage at the cathode would have given the same
result. In fact, Sarkar et al also observed a transient effect when the voltage was
disconnected for 30 s and the experiment continued afterward: The increase
in resistance had disappeared while the deposit was still on the electrode. It
was only after about 30 s that the potential required for the constant applied
current reached the value it had reached before the experiment was interrupted.
Such a transient effect is more easily understood if the increase in potential
drop is caused by concentration overvoltage. During the time the experiment
is paused, the formed concentration gradients can be relieved by diffusion, and
the concentration overvoltage needs to be built up once again.
It should be noted that a higher specific resistance of the deposit compared
with the suspension is indeed expected. In a deposit, the volume fraction of
the liquid is lower than in the suspension, and therefore the effective area for
ionic conductance is lower (17). Pierce (3) found a high resistance for dense
layers of a cathodically applied polymeric primer. If the deposit was a porous
film, however, as sometimes observed during the electrodeposition of latex, the
voltage drop over the film remained very low due to the conductance through
the pores in the film (3). Since during deposition of ceramic or metal powders
the volume fraction of solids in the deposit is typically between 40 and 60
vol%, the resistance of the deposit cannot be many times higher than the ionic
resistance of the suspension.
The resistance of the suspension is also not necessarily constant. Vandeperre
& Van der Biest (93) modeled the conductivity of a suspension as consist-
ing of a contribution of ions outside the double layers of the powder particles
and a contribution by the electric charges associated with the moving particles
(Figure 7). The decrease in current during electrophoretic deposition of SiC
from acetone with addition of N-butylamine was found to be mainly from a de-
crease in suspension conductivity (93) caused by depletion of the suspension.
Naturally, the decrease in conductivity of the suspension through this mech-
anism is observed only if the amount of powder deposited is large compared
with the total amount of powder in suspension and if the ionic conductivity is
relatively small. Moreover, another factor contributing to the lower conductiv-
ity of the deposit compared with the suspension is that while in the suspension
the movement of the particles contributes to the conductivity, in the deposit the
particles are immobile and therefore do not contribute to the conductivity of the
deposit. Figure 8 shows the evolution of the current and yield as a function of
time for deposition of alumina (Baikowski-type SM8, 100 g · L−1) from tech-
nical ethanol with addition of 1 M acetic acid (72), together with the calculated
current and yield based on the description of the conductivity of suspensions as
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
Figure 7 Schematic diagram for the resistance of the deposition cell. In the suspension, moving
ions as well as moving particles contribute to the conductivity, which can be represented by two
resistors (Rpowder & Rion) in parallel. In the deposit, only ions are moving, hence one resistance,
Rdep, in series with the resistance of the suspension. Finally, there is also a potential drop, 1i , at
each electrode. The value of the resistance related to the contribution of the powder is a function
of the solid concentration.
Figure 8 Measured evolution of current (filled symbols) and yield (open symbols) during elec-
trophoretic deposition of alumina from technical ethanol with 1 M acetic acid. The solid lines
were calculated based on the assumption that the decrease in current is caused by the vanishing
contribution of the alumina particles to the conductivity of the suspension, whereas the dotted lines
were calculated assuming that the current decrease was the result of very high specific resistance of
the deposit and no contribution of the powder to the conductivity. Although the calculated currents
coincide, the evolution of the yield is better described by the assumption that the conductivity of
the suspension decreases.
If, however, the volume fraction of solids is low, the particles can deposit at a
rate proportional to their individual electrophoretic mobility (37, 38, 47, 95). In
work by Moritz on deposition of aluminum nitride with incorporation of yttria
as a sintering additive, an equal deposition rate was observed from solids load-
ing higher than 28 vol% (95). These observations are in accordance with the
findings by Massoud (25), who states that in slips for sanitary ware no segre-
gation between the components was found. The suspension used in his work
contained a high volume fraction of solids.
CONCLUSIONS
EPD of any material available in the form of a fine powder or colloidal suspen-
sion is possible. The technique allows the application of coatings, the shaping
of objects, and the infiltration of porous structures and woven fiber preforms.
Moreover, laminated and graded materials can also be made. It is a fast and
economical process that can be automated for use on an industrial scale. The
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
Literature Cited
1. Heavens SN. 1990. In Advanced Ceramic 2. Hiemenz PC. 1986. Principles of Col-
Processing and Technology, ed. JGP Bin- loid and Surface Chemistry. New York:
ner, 7:255–83. Park Ridge, NJ: Noyes Dekker. 815 pp. 2nd. ed.
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
3. Pierce PE. 1981. J. Coat. Techn. 53(672): 27. Aveline M. 1966. Ind. Céram. 581:28–
52–67 31
4. Eberhard J. 1997. Galvano–Organo 28. Boncoeur M, Carpenter S. 1972. Ind. Ce-
66(678):621–23 ram. 648:79–81
5. Lamb VA, Reid WE. 1960. Plating 29. Chronberg MS, Händle F. 1978. Interce-
(3):291–96 ram 1:33–34
6. Engelhardt Th, Rijenen P, Hellkuhl L, 30. Andrews JM, Dracass J, Collins AH, Cor-
Kaup F, Warnke H. 1990. Mitt. Ver. Dtsch. nish DC. 1969. Proc. Brit. Ceram. Soc.
Emailfachleute 38(4):37–48 12:211–29
7. Engelhardt Th, Rijenen P, Hellkuhl L, 31. Nass R, Schmidt H, Storch W, Harbach F,
Kaup F, Warnke H. 1990. Mitt. Ver. Dtsch. Neeff R, Nienburg H. 1992. In Ceramic
Emailfachleute 38(5):60–74 Powder Processing Science, ed. H Haus-
8. Engelhardt Th, Rijenen P, Hellkuhl L, ner, GL Messing, S Hirano, pp. 625–32.
Kaup F, Warnke H. 1990. Mitt. Ver. Dtsch. Heidelberg: Bovery
Emailfachleute 38(6):86–91 32. Jean JH. 1995. Mater. Chem. Phys. 40:
9. Engelhardt Th, Rijenen P, Hellkuhl L, 285–90
Kaup F, Warnke H. 1990. Mitt. Ver. Dtsch. 33. Ferrari B, Farinas JC, Moreno R. 1998.
Emailfachleute 38(7):86–91 Presented at World Ceramics Congress
10. Engelhardt Th, Rijenen P, Hellkuhl L, and Forum on New Materials, Florence,
Kaup F, Warnke H. 1990. Mitt. Ver. Dtsch. Italy
Emailfachleute, 38(8):105–16 34. Vandeperre L, Van der Biest O, Bouyer F,
11. Engelhardt Th, Rijenen P, Hellkuhl L, Persello J, Foissy A. 1997. J. Eur. Ceram.
Kaup F, Warnke H. 1990. Mitt. Ver. Dtsch. Soc. 17(2):373–76
Emailfachleute, 38(9):120–124 35. Vandeperre LJ, Van der Biest O, Bouyer
12. Ortner M. 1964. Plating (9):885–89 F, Foissy A. 1998. Ceram. Bull. 77(1):53–
13. Barraclough M, Bolton NE, Collins AH, 58
Andrews JH. 1967. IEEE Trans. Magnet. 36. Wittwer H, Krüger HG. 1995. Ber DKG
3(3):531–34 72(9):556–60
14. Shane MJ, Talbot JB, Kinney BG, Sluzky 37. Moritz K, Reetz T. 1993. Ber DKG
E, Hesse KR. 1994. J. Colloid Int. Sci. 70(7):348–50
165:334–40 38. Moritz K, Reetz T. 1993. In Third Eu-
15. Shane MJ, Talbot JB, Schreiber RD, Ross roceramics, ed. P Duran, JF Fernandez,
CL, Sluzky E, Hesse KR. 1994. J. Colloid 1:425–30. Castellón de la Plana, Spain:
Int. Sci. 165:325–33 Faenza Editrice
16. Siracuse JJ, Talbot JB, Sluzky E, Hesse 39. Kennedy JH, Foissy A. 1975. J. Elec-
KR. 1990. J. Electrochem. Soc. 137(1): trochem. Soc. 122(4):482–86
346–48 40. Powers RW. 1986. Ceram. Bull. 65(9):
17. Sussman A, Ward TJ. 1981. RCA Rev. 1270–77
42(6):178–97
18. Miziguchi J, Matsumura M, Suzuki M, 41. Nicholson PS, Sarkar P, Huang X. 1993.
Yamato H. 1991. J. Electrochem. Soc. J. Mater. Sci. 28:6274–78
138(10):2942–46 42. Bissinger M, Prakash O, Sarkar P, Nichol-
19. Nojima H, Koba M, Shintaku H, Na- son PS. 1994. Ceram. Sci. Eng. Proc.
gata M. 1991. Jpn. J. Appl. Phys. 30(7): 15:1084–92
L1166–68 43. Ferrari B, Sanchez–Herencia AJ, Moreno
20. Hein M, Peiniger M, Mahner E, Müller G, R. 1998. Mater. Lett. 35:370–74
Piel H, et al. 1989. Physica C 162:111–12 44. Fischer R, Fischer E, De Portu G, Roncari
21. Hein M, Peiniger M, Piel H, Ponto L, E. 1995. J. Mater. Sci. Lett. 14:25–27
Becks M, Klein U. 1989. J. Appl. Phys. 45. Vandeperre L, Van der Biest O, Clegg
66(12):5940–43 WJ. 1997. In Key Engineering Materials,
22. Chu CT, Dunn D. 1989. Appl. Phys. Lett. 127:567–74
55(5):492–94 46. Vandeperre L, Van der Biest O. 1998. Sil-
23. Sharma AD, Sen A, Maiti HS. 1993. Ce- icates Ind. 62(5):87–91
ram. Int. 19:65–70 47. Vandeperre L, Van der Biest O. 1998. SiC
24. Woolf LD, Fagaly RL, Raggio WA, Elsner laminates with porous interlayers pro-
FE, Fisher MV, et al. 1991. Appl. Phys. duced by EPD. Presented at 100th Ann.
Lett. 58(5):534–36 Meet. Am. Ceram. Soc. Cincinnati, OH
25. Massoud E. 1979. Interceram 2:117–19 48. Sarkar P, Nicholson PS, Huang X. 1993.
26. Mihailescu M, Emandi M, Vancea V, J. Am. Ceram. Soc. 76(4):1055–56
Marcu M. 1991. Interceram 40(3):165– 49. Sarkar P, Datta S, Nicholson PS. 1997.
70 Composites Part B 28:49–56
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
50. Zhao C. Vandeperre L, Vleugels J, Van 76. Blanco Lopez MC, Rand B, Riley F.
der Biest O. 1998. J. Mater. Sci. Lett. 1996. Key Eng. Mater. 132–136:305–
17:1453–55 8
51. Nicholson PS, Sarkar P, Datta S. 1996. 77. Hamaker HC, Verwey EJW. 1940. Trans.
Ceram. Bull. 75(11):48–51 Farad. Soc. 36:180–85
52. Scala LC, Sandor JE. 1968. Electrochem. 78. Koelmans H. 1955. Philips Res. Rep.
Technol. 6(11–12):434–37 10:161–93
53. Gal-Or L, Haber S, Liubovich S. 1992. J. 79. Grillon F, Fayeulle D, Jeandin M. 1992.
Electrochem. Soc. 139(4):1078–81 J. Mater. Sci. Lett. 11:272–75
54. Ishihara T, Takita Y, Sato K. 1996. J. Am. 80. Sarkar P, Nicholson PS. 1996. J. Am. Ce-
Ceram.Soc. 79(4):913–19 ram. Soc. 79(8):1987–2002
55. Streckert HH, Norton KP, Katz JD, Freim 81. Russ BE, Talbot JB. 1998. J. Electrochem.
JO. 1997. J. Mater. Sci. 32:6429–33 Soc. 145(4):1253–56
56. Kooner S. 1995. Mater. Challenge (8):27- 82. Bouyer F, Foissy A. 1999. J. Am. Ceram.
29 Soc. In press
57. Kawai C, Wakamatsu S. 1996. J. Mater. 83. Brown DR, Salt FW. 1965. J. Appl. Chem.
Sci. 31:2165–70 15(1):40–48
58. Boccaccini AR, Ponton CB, MacLaren 84. Sarkar P, Nicholson PS, Prakash O,
I, Lewis MH. 1997. J. Eur. Ceram. Soc. Wang G. 1994. Ceram. Sci. Eng. Proc.
17:1545–50 15(5):1019–27
59. Boccaccini AR, Trusty PA. 1998. J. 85. Sheppard SE. 1927. Trans. Am. Elec-
Mater. Sci. 33:933–38 trochem. Soc. 52:47–82
60. James RO, Davis JA, Leckie JO. 1978. J. 86. Sarkar P, Huang X, Nicholson PS. 1993.
Colloid Int. Sci. 65(2):331–44 Ceram. Eng. Sci. Proc. 14:707–17
61. Lyklema J. 1984. J. Colloid Int. Sci. 99(1): 87. Hamaker HC. 1940. Trans. Farad. Soc.
109–17 36:279–87
62. Ferrari B, Moreno R. 1997. J. Eur. Ceram. 88. Gutierrez CP, Wallace TC, Mosley JR.
Soc. 17:549–56 1962. J. Electrochem. Soc. 109:923–
63. Wang G, Nicholson PS, Sarkar P. 1997. J. 27
Am. Ceram. Soc. 80(4):965–72 89. Sarkar P, Nicholson PS. 1995. J. Am. Ce-
64. Powers RW. 1975. J. Electrochem. Soc. ram. Soc. 78(11):3165–66
90. Zhang Z. 1995. J. Am. Ceram. Soc.
122(4):482–86 78(11):3167–68
65. Fisch HA. 1972. J. Electrochem. Soc. 91. Dunbar JC, Mitchell TJ. 1953. J. Sci. Food
119(1):57–64 Agr. 4:190–96
66. Fessenden RJ, Fessenden JS. 1986. Or- 92. Ferrari B, Moreno R. 1995. In Fourth
ganic Chemistry. Monterey, CA: Brooks- EuroCeramics, ed. C. Galassi, 1:421–
Cole. 3rd ed. 28. Faenza, Italy. Gruppo Editoriale
67. Labib ME, Williams R. 1986. Colloid Faenza Editrice
Polym. Sc. 264:533–41 93. Vandeperre LJ, Van der Biest OO. 1998.
68. Labib ME, Williams R. 1984. J. Colloid In Innovative Processing and Synthesis of
Int. Sci. 97(2):356–66 Ceramics, Glasses, and Composites, Ce-
69. Jensen WB. 1978. Chem. Rev. 78(1):1– ram. Proc., ed. NP Bansal, KV Logan, JP
21 Singh, Ceram. Trans. Vol. 85. Westerville
70. Okuyama M, Haggerty JS, Garvey G, OH: Am. Ceram. Soc.
Ring TA. 1989. J. Am. Ceram. Soc. 94. Bockris JOM, Reddy AKN. 1973. In
72(10):1918–24 Modern Electrochemistry, New York:
71. Bolger JC. 1981. In Acid–Base Inter- Plenum Rosseta. 1432 pp. Vol. 1 & 2
actions, ed. KL Mitchel. New York: 95. Moritz K, Müller E. 1997. Werkstoff–
Elsevier und Verfahrenstechnik (DGM) 563–
72. Vandeperre L. 1998. Electrophoretic 68
shaping for ceramic laminated compos- 96. Kuwabara K, Noda Y. 1993. J. Mater. Sc.
ites. PhD thesis, Katholieke Univ., Leu- Lett. 28:5257–61
ven, Belgium. 220 pp. 97. Mizuguchi J, Muchi T, Sumi K. 1983. J.
73. Vandeperre L, Van der Biest O. 1999. J. Electrochem. Soc. 130(9):1819–25
Eur. Ceram. Sci. Submitted 98. Wakamatsu KS. 1995. J. Mater. Sci. Lett.
74. Vandeperre L, Zhao C, Van der Biest 14:467–69
O. 1999. 6th Conf. Eur. Ceram. Soc. 99. Furman VV, Gaiduchenko GK, Vlastuk
Brighton, UK. In press RZ, Deimontovich VB. 1983. Porosh.
75. Kennedy JH, Foissy A. 1977 J. Am. Ce- Metall. 251(11):83–87
ram. Soc. 60(1):33–36 100. Valdes JL, Huggins H, Mitchel, JW,
P1: APR/SPD P2: APR/ARY QC: APR
May 15, 1999 14:52 Annual Reviews AR086-11
Mucha JA, Seibles L. 1991. J. Elec- 108. Caley WF, Flengas SN. 1976. Can. Met.
trochem. Soc. 138(2):635–36 Quart. 15(4):375–82
101. Zhitomirsky I. 1998. Mater. Lett. 37(9): 109. Hang KW, Anderson WM, Andrus J.
72–78 1981. RCA Rev. 42(6):159–77
102. Mahapatra AK, Dhananjayan N. 1981. 110. Baumgartner CE, Grimaldi JJ, DeCarlo
Trans. Ind. Inst. Metals, 34(6):495–500 VJ, Glugla PG. 1985. J. Electrochem. Soc.
103. Hill CGA, Rees ALG, Lovering PE. 1947. 132(1):57–63
Trans. Farad. Soc. 43:407–17 111. Hirayama C, Berg D. 1963. Electrochem.
104. Pearlstein F, Gallacio A, Wick R. Technol. 1(7–8):224–27
1963. J. Electrochem. Soc. 110(7):843– 112. Hosseinbabaei F, Raissidehkordi B. 1998.
46 Proc. 9th CIMTEC Conf. TECHNA Srl,
105. Giersig M, Mulvaney P. 1993. J. Phys. Faenza, Italy. In press
Chem. 97(24):6334–36 113. Grosso PF, Rutherford RE, Sargent DE.
106. Standish JV, Boerio FJ. 1980. J. Coat. 1970. J. Electrochem. Soc. 117(11):1456–
Technol. 52(663):29–39 59
107. Nagai M, Takuma Y, Yamashita K, Ume- 114. Williams EW, Jones K, Griffiths J, Rough-
gaki T. 1993. J. Am. Ceram. Soc. 76(1): ley DJ, Bell JM, et al. 1979. Solar Cells
253–55 1:357–66