You are on page 1of 20

Fundamentals of Electrodeposition

G Zangari, University of Virginia, Charlottesville, VA, USA


© 2018 Elsevier Inc. All rights reserved.

Definition and General Description of the Electrodeposition Process 141


A Brief History: From Gilding and Heavy Machinery to Nanotechnology 141
Implementation of Electrodeposition Processes 142
Electrochemistry Fundamentals: Charge and Mass, Thermodynamics, Kinetics, and Interface Structure 143
Stoichiometry 143
Thermodynamics 143
Kinetics 144
Structure of the Electrode/Electrolyte Interface 145
Elementary Steps in Electrodeposition 146
Transport of ions and current distribution at various length scales 146
Charge transfer mechanism and surface diffusion 148
Incorporation in the Growing Film 148
Metal Deposition: Rationalization, Trends, and Some Examples 150
Copper 150
Iron Group Metals (Nickel, Cobalt, Iron) 151
Zinc 152
Alloy Deposition 152
Metal Oxides and Chalcogenides 155
Electrodeposition of Oxides 156
Electrodeposition of Chalcogenides 157
Evolution of Film Structure and the Control over Morphology 158
Electrodeposition of Composite Materials 158
Nanostructured Materials 159
References 160

Definition and General Description of the Electrodeposition Process

Electrodeposition is a material production technology using an applied potential or current flowing through an electrolytic solution
that contains metal ions. These metal ions are eventually reduced to atoms in an electrically conducting substrate, thus forming
metals, alloys, metal-based compounds, or composites as a coating or in form of powders. This technology has a long history, start-
ing at the beginning of the 19th century, when the first constant voltage source, the Volta pile, became available. The formation of
metallic coatings by metal ion reduction, however, had been practiced much earlier, starting from the 2nd–3rd century, when this
task was carried out by using natural acids to dissolve metals to form ions and subsequently reduce them on the surface of interest.
Initially, electrodeposition was used for decorative applications, and only later for engineering purposes. Through the industrial
revolution and, about one century later, the digital revolution, the nature of this technology has significantly changed, and today,
while it is still used for conventional engineering applications, it has gained an important role as an efficient and versatile micro-
and nanofabrication method for a variety of advanced applications.
This article summarizes the history of this technology and introduces the fundamentals of this discipline at the macroscopic level
and then the atomistic processes underlying electrochemical material growth. The correlation between operational parameters and
the characteristics of the materials will be discussed, and the various classes of materials that can be obtained will be described,
including metals, alloys, metal oxides and chalcogenides, composites, and nanomaterials.

A Brief History: From Gilding and Heavy Machinery to Nanotechnology

Electrodeposition was invented shortly after the Volta pile, the first constant voltage source, was developed in 1800.1 Three years
later Brugnatelli, using the Volta pile, was able to plate gold on “two large silver medals,” but after that, limited development
was reported until 1830. The advent of the Daniell cell (1836) provided a more reliable source of current which could be utilized
for the formation of thin and continuous coatings. This opened the way to the patenting and commercial development of the
gilding process by the Elkingtons in Birmingham, UK (1840), with parallel discoveries happening in Russia as well. Around the

141
142 Fundamentals of Electrodeposition

1850s, practical electrolyte formulations for Ni, Zn, and Sn plating among others were also developed, opening the way to engi-
neering applications of coatings for both decorative and protective uses.2
In the early days of electroplating, this process was indeed used only for decorative applications; even today many electroplated
articles are coated for such purposes, using Au, Ag, and their alloys to embellish precious items such as art pieces or jewelry. Chro-
mium is also an important decorative coating, as it combines a pleasant appearance with outstanding corrosion and abrasion resis-
tance. What is called a Cr coating is often a three-layer stack of electroplated Cu/Ni/Cr, where Cu insures adhesion to the substrate,
Ni provides mechanical strength, and Cr abrasion resistance and appearance. Many electroplated coatings indeed are applied to
provide both mechanical protection and corrosion resistance to a mechanically strong bulk; in other instances, the plated material
could be used instead to repair worn parts.3
Corrosion-resistant coatings, in particular, may protect the bulk material from the environment or could provide sacrificial
protection by corroding preferentially. Zinc works very well for this purpose, and its operating life can be significantly increased
by alloying it with Co or Ni, or by depositing an additional coating such as a phosphate compound. Cadmium in this context works
even better, but due to its toxicity, applications of electrodeposited Cd are limited and tend to decrease over time.
More specialized applications include Cu for electrical conductivity, alloys of Au with transition metals (TMs; Ni, Co, Pd) to
provide low contact resistance in printed circuit boards (PCBs), Pt coatings for extreme environments such as high-temperature
turbines, Pb and metal sulfides for soft bearings. Ni and Cu are also extensively utilized in electroforming and microsystems as struc-
tural materials. Synthesis and applications of electroplated materials are discussed in detail in Schlesinger and Paunovic.4
Starting with the 1950s, electrochemical processes progressed beyond the realm of engineering applications and expanded to
electronic manufacturing.5 In electronics, miniaturized silicon-based devices were initially manufactured on a surface by a combi-
nation of lithography, which is used to define regions where metals are to be deposited, and film deposition. The metallic films were
generally grown by vapor phase methods and did form blanket films, the excess parts of which needed to be removed. Electrode-
position, in contrast, was capable to deposit only at conductive regions, therefore enabling for the first time purely additive growth
and closer reproduction of the lithographic pattern. Starting with low-end PCBs, this technique demonstrated the ability to provide
the needed precision while allowing fast production and low material waste. Over time, new processes and capabilities were devel-
oped, enabling manufacturing of more complex devices with smaller features. Interconnects between the semiconductor chip and
the packaging were made first by wire bonding and later replaced by solder pads, which evolved to multichip-modules intercon-
nections. These conductive patterns include surface electrical interconnects, as well as vias between different levels through the
formation of through-hole contacts. At a smaller scale, interconnects among transistors within the semiconductor chip were
made by vapor deposition of Al wires, later displaced by electroplated Cu lines, which provided higher precision, improved conduc-
tivity, and lower electromigration.6 This so-called Damascene process, from the city of Damascus where a similar process was devel-
oped to form precious metal inlays in swords and tools, achieved outstanding prominence as it has been capable to continue the
rate of miniaturization of the smallest features down to pitches of 22 nm or less.
Electrodeposition through lithographic masks was also applied to the growth of magnetic materials, enabling significant mini-
aturization of write/read heads in magnetic recording systems, thus sustaining the ongoing exponential increase of recording density
in magnetic storage since the 1980s.7
The latest steps in the evolution of electroplating applications include (i) the production of micron-sized parts by electroforming
in three dimensions, competing with more complex additive manufacturing methods,8 and (ii) the inclusion of sensing and actu-
ating capabilities in microdevices, integrating the electrical functionality with a variety of other functions, including biomedical
ones.

Implementation of Electrodeposition Processes

The electrochemical growth of metallic films occurs in a container, or tank, with two electrodes immersed in the electrolyte, each
being in contact to the same external electrical power source; electrodeposition occurs by applying a current through the solution
between the two electrodes, resulting in film growth at the negatively polarized electrode, the cathode, where the metal ion reduc-
tion process occurs (Fig. 1). In order to close the circuit, an oxidation process must occur at the opposite electrode, the anode; this
reaction may be either dissolution of the anode or oxidation of an electrolyte species, in some cases water. Laboratory cells utilize in
addition a third electrode to closely monitor and control the potential of the cathode with respect to a reference electrode. This
electrode ideally maintains a constant potential, irrespective of the magnitude of the current flowing between anode and cathode,
by running a highly reversible electrochemical reaction. All commercial plating cells instead utilize only two electrodes and current
or cell voltage control, in order to simplify operation. Typical laboratory systems use cells with minute volumes, varying approxi-
mately from 1 mL to 1 L; commercial tanks, in contrast, can be quite large, of the order of meters in size. The tank should not react
with the electrolyte; for an alkaline solution a mild steel liner could be used, while for an acidic electrolyte, rubber or plastic should
be employed. Specialty tanks are used when many small items must be electroplated at the same time; in this case, the items may be
hung up to a rack, or may instead move within a rotating insulated vessel, whereby contact is made by suitable cables. Finally,
continuous plating of long metal strips could be implemented by reel-to-reel plating in plants tens of meters long.
The plating electrolyte is obtained by dissolving metal-containing salts in a solvent, usually water; under these conditions, the
ions become hydrated or are coordinated with chelating agents, mostly anions, to help maintain the metal ions in solution in a wide
range of conditions. An unreactive salt is usually added to increase conductivity without reacting with the electrodes. A variety of
Fundamentals of Electrodeposition 143

Fig. 1 A typical electrochemical cell set-up for electroplating Copper using a soluble anode. Figure from https://en.wikipedia.org/wiki/Electroplating.

neutral organic molecules adsorbing selectively at definite crystallographic facets are used to control growth morphology and crys-
tallography, while surfactants are employed to control the surface energy at the substrate, thus affecting growth modes.

Electrochemistry Fundamentals: Charge and Mass, Thermodynamics, Kinetics, and Interface Structure9
Stoichiometry
Electrodeposition as an electrochemical process must follow the laws of electrochemistry. Every electrochemical reaction does not
only need to satisfy the mass balance but also the balance of charge. Faraday, in 1833, formulated the two laws dictating mass and
charge balance in electrochemical reactions. The first law states that the number of moles n produced or consumed in an electro-
chemical reaction must be proportional to the charge flowing through the circuit, Q. The second law states that the mass W of
different substances formed by the flow of the same charge is proportional to the molar mass A divided by the oxidation state
of the species, z. In mathematical form, the two laws can be summarized by the expression:
W ¼ A$Q=zF (1)
where Q ¼ I$ t is the total charge (current  time) flowing through the electrochemical cell during the process of interest, and the
Faraday constant F ¼ 95.485 C/mol ¼ 26.8 A h/mol corresponds to the charge of one mole of electrons.
Since the film is deposited on a substrate of assigned area, Eq. (1) can also be written per unit area:
W ¼ A$I$t=S$z$F ¼ A$j$t=zf (2)
where S is the area of the substrate and j ¼ I/S is the current density (CD). From Eq. (2) it is also possible to derive an equation
for the film thickness d generated by a CD applied for a given time:
d ¼ A$j$t=z$F$r (3)
where r is the material density. Note that Eqs. (1), (2) and (3) assumes that all the current through the electrochemical cell is
used to grow the material of interest, without parasitic reactions.

Thermodynamics
When a metal Me is immersed in a solution containing the corresponding ion Mez þ, a dynamic equilibrium is achieved over time
where the rate of metal dissolution from the electrode equals the rate of metal ion reduction:
Mezþ þ ze 5Me (4)
This equilibrium condition is characterized by a charge separation and therefore a potential difference at the metal/electrolyte
interface. This potential difference can only be measured with respect to another electrochemical reaction that reaches equilibrium
very quickly and is stable over time, working as a reference electrode. One of such reactions is the hydrogen reduction/oxidation
(redox):
2Hþ þ 2e 5H2 (5)
144 Fundamentals of Electrodeposition

Once this reaction is arbitrarily assigned a redox potential E(Hþ/H2) ¼ 0 under standard conditions (standard hydrogen elec-
trode, SHE), the equilibrium potential of any other redox reaction can be defined operationally by experimentally measuring
the potential difference in a cell consisting of the hydrogen electrode and the electrode under study. More information on reference
electrodes is available in Inzelt et al.10
Under nonstandard states, the conditions of equilibrium change due to the activity of the ionic species, a quantity related to the
ability of the system to perform work; activity approximates the concentration of the species in solution at concentration of the
order of 10 3 mol/L and below. Under these conditions, the expression for the redox potential E(Mez þ/Me) becomes:
 
E ¼ E0 þ ðRT=zF Þln Mezþ (6)

where E0 is the standard redox potential, R is the gas constant (8.314 J/mol K), T the absolute temperature, “ln” stands for the
natural logarithm, and [Mez þ] is the concentration of the metal ion in solution (mol/L).

Kinetics
When the equilibrium is disturbed by applying an external potential, the metal ions will be reduced at the substrate (the applied
potential is more negative than E) or the metal will be oxidized to metal ions (the applied potential is more positive than E). Note
that in the latter case, the metal may be either dissolved as ions or form a metal oxide; this depends on the electrolyte chemistry. A
Pourbaix diagram can be consulted to determine the regions of stability of the metal, metal oxide, or metal ions in a plot of potential
vs. pH.
In practice, however, upon applying an external potential to the electrode of interest the system under study may either be inert
within a small range of applied potentials (polarizable electrode) or may start to flow current immediately (polarizable interface),
resulting in Faradaic processes leading to electrochemical transformation. The driving force for this transformation is the overpo-
tential, defined as
h ¼ V  Eeq (7)

where V is the applied voltage and Eeq the equilibrium potential.


Note that a negative overpotential results in metal reduction and electrodeposition. The mechanism of the metal ion reduction
to a metal at the substrate consists of a series of elementary steps pictured in Fig. 2. The metal ions diffuse towards the electrode,
possibly undergoing a chemical or structural transformation, then intermediate species adsorb at the electrode, where they are
reduced and finally incorporated in the growing film.
Each elementary step is characterized by its own reaction rate; if one of them, E, is by far the slowest step, the overall rate of the
process is dictated by step E, called the rate-determining step (RDS). If this condition is not satisfied, it is still possible to determine
the overall reaction rate by assuming steady state and imposing that (i) each step has the same rate and (ii) the concentration of each
intermediate is constant over time.
At low overpotential, the driving force for electron transfer (ET) is small and therefore this is the slowest step; the rate expression
for the corresponding one-electron reaction: Meþ þ e 5 Me is determined by the Marcus theory.11 The ET theory in a homoge-
neous medium for a reaction Ox þ e 5 Red (Ox, Red being a generic oxidized and reduced species, respectively) was first devel-
oped by Marcus, for which he earned the Nobel prize in 1992. In this theory, the ET is described by assuming that each ion acts as an
oscillator (i.e., the point-like entity is subjected to elastic interactions while vibrating) with various energy levels. ET may occur when
the vibrational energy is larger than the energy barrier DGz (Fig. 3), and its probability of occurrence P is calculated under the condi-
tion that the configuration of donor and acceptor remain unchanged during ET; P depends on the energy barrier DGz and the

Fig. 2 General mechanism of metal ion reduction. The metal ion Mezþ diffuses to the electrode/electrolyte interface, and may undergo chemical
reactions, adsorption and surface diffusion before being reduced by one or more electron transfer steps. Copyright G. Zangari.
Fundamentals of Electrodeposition 145

Fig. 3 Energy landscape for homogeneous electron transfer. The energy of reactants (R) and products (P) are modeled as ideal oscillators, DG0 is
the difference in the free energy minimum of P and R, and l is the reorganization energy, i.e. the energy required to induce structural changes that
accomplish the optimum configuration to enable electron transfer.

reorganization energy l (Fig. 3). The theory of ET has been later extended to take into account the actual electronic structure of the
electrode and the possibility of electron tunneling, which depends exponentially on the distance between the reducing ion and
electrode.
By considering the relative shift of the potential landscapes for the Mez þ (reactant) and Me (product) species due to the ET it is
possible to write the energy barrier as function of the applied voltage and therefore the rate of ET as a function of the overvoltage h,
the so-called Butler–Volmer (BV) equation:
j ¼ j0 ½expðaf hÞ  expð ð1  aÞf hÞ (8)
where j is the CD, j0 the exchange CD, a is the transfer coefficient (a measure of the symmetry of the energy barrier for oxidation
vs. reduction), f ¼ F/RT.9 This expression is valid only for single ET; most electrode processes, however, consist of several steps,
including chemical, electrochemical, adsorption processes, and electrochemical reactions involving the electrode. The exchange
of more than one electron should rigorously be derived assuming a probable mechanism and steady state conditions, but instead
it is often described by the same BV equation modified to take into account the transfer of z electrons. While the probability of such
a process is very small, it is found experimentally that in many cases, the reduction of divalent metal ions seem to follow a pseudo-
BV equation where the number of electrons exchanged is included in the equation:
j ¼ j0 ½expðzaf hÞ  expðzð1  aÞf hÞ (9)
when the applied overpotential is sufficiently large, of the order of 120 mV, one of the two exponential can be neglected. Consid-
ering cathodic applied voltages and electrodeposition, the BV equation simplifies to
j ¼ j0 ½expðzaf hÞ (10)
which is often reframed in the logarithmic form:
 
h ¼ ðRT=zaF Þln j j0 (11)

since this equation was used for the first time by Tafel to describe experimental data in electrochemical kinetics.

Structure of the Electrode/Electrolyte Interface


As discussed earlier, if a Mez þ/Me redox reaction with redox potential E(Mez þ/Me) can occur at the electrode, the reaction usually
does not start exactly at the redox potential, instead, a finite overvoltage honset is needed. For any |h| < |honset |, no Faradaic reaction
occurs, but the microscopic charge distribution at the electrode/electrolyte interface changes, resulting in a capacitance at this inter-
face (Fig. 4). The distribution of ions at the electrolyte interface side must balance the charge at the electrode surface and consists of
specifically adsorbed ions at the inner (IHP) and outer (OHP) Helmholtz plane, as well as a diffuse distribution of ions due to the
balance between electrostatic interactions and thermal motion. This double layer generates a localized voltage drop of  1 V and
extremely high electric fields in a thickness (the Debye length) which varies between less than a nanometer to several 100 nm.
The effective double layer thickness L depends among others on the overall metal ion concentration: at high concentration L is
small, at low concentration L is larger. A better description of the double layer has been proposed, whereby the ions close to the
electrode and those distributed in the electrolyte form two capacitors in series; this model predicts the experimentally measured
behavior of the interface capacitance as a function of ion concentration and applied voltage. The presence of water, anions, and
146 Fundamentals of Electrodeposition

Fig. 4 The equilibrium distribution of charges at an electrolyte/metallic electrode interface. Taken from Wu, J.; Risalvato, F. G.; Ke, F.-S.; Pellechia,
P. J.; Zhou, X.-D. Electrochemical Reduction of Carbon Dioxide I. Effects of the Electrolyte on the Selectivity and Activity with Sn Electrode.
J. Electrochem. Soc. 2012, 159(7), F353–F359.

other molecules that may adsorb at the electrode further complicate the description of the interface and therefore of the deposition
process.12

Elementary Steps in Electrodeposition

Electrodeposition occurs through a series of elementary steps starting from the metal ions solvated in the electrolyte and ending in
the incorporation of the metal atom in the crystal structure of the coating being formed. In the following, each step and the physical
locations where they occur will be discussed in detail (Fig. 5).

Transport of ions and current distribution at various length scales


Metal ions in the bulk electrolyte are bonded to water molecules by ion–dipole forces, or bound to complexing molecules via dipole
or covalent interactions. In the presence of an applied potential, mass transfer of the ions towards the cathode occurs via three
distinct processes: (i) convection (macroscopic motion of the fluid due to stirring, density gradients originating eddies, or hydro-
dynamic transport), (ii) diffusion (ion motion due to gradients in chemical potential, most often concentration gradients), and (iii)
migration (motion of the ions due to an applied electric field). Accordingly, the mass flux Fi (mole/s m2) of a given species i is given
by the Nernst–Planck equation, which includes all three processes in the above order and in one dimension reads:
Fi ðxÞ ¼ Ci $vðxÞ  Di dCi =dx  mi Ci dV=dx (12)

Fig. 5 Schematics of the various elementary steps involved in electrodeposition and the physical locations where they occur. The hydrated ion
initially flows due to convection processes until it enters the diffusion layer, where it travels only by diffusion; close to the electrode/electrolyte inter-
face, the hydration sheath is initially aligned and then removed by the large electric field, enabling charge transfer and incorporation in the growing
film. Copyright Giovanni Zangari.
Fundamentals of Electrodeposition 147

Ci being the concentration of i (mole/m3), v(x) the velocity of a volume element, Di is the diffusion coefficient (m 2/s), dC/dx
the concentration gradient, mi the mobility (velocity per unit electric field), and dV/dx the potential gradient. Ion motion in the
bulk of the electrolyte is generally dominated by convection, either natural (uncontrolled) or forced (controlled) by purposeful
mechanical forces, obtained for example by using a rotating disk electrode (RDE), for which the current distribution could be
calculated exactly and results ideally in a uniform current distribution. Closer to the electrode ( mm), in the so-called hydro-
dynamic layer, convection phenomena decay due to mechanical friction, and transport is due to diffusion and migration only. In
general, quantifying the mass transport in electrochemical systems is complicated by the fact that ion motion induces changes in
the electric field distribution which in turn alters the forces acting on the ions, such that a self-consistent solution for the mass flux
is difficult to obtain. In order to avoid these complications, a high concentration of an electrochemically inert, unreacting salt is
added to the electrolyte to minimize the effect of the electric field while enhancing the influence of the concentration gradient. In
other words, the resulting high electrical conductivity minimizes electric fieldsdand therefore migrationdin the electrolyte bulk.
Under these conditions, the ion transport is overwhelmingly due to the concentration gradient generated by the electrochemical
process occurring at the electrode; therefore the ionic motion is in the same direction irrespective of their charge or lack thereof.
The mass flux can therefore be simplified to the form: Fi ¼  Di dC(x)/dx, calculated at the electrode position (x ¼ 0), while the
current is similarly written as
Ji ¼ zF$Di dCðxÞ=dx at x ¼ 0 (13)
The thickness of the region where Eq. (13) holds may vary from 0.1 mm for unstirred solutions to tens of micrometers under
forced stirring. For non-steady state mass transfer, the concentration of the reacting species tends to change over time, resulting in
a time-dependent diffusion process, ideally described by the following equation:

jdiff ¼ zFCN ðD=pt Þ1=2 (14)


where CN is the ion concentration in the electrolyte bulk.
The local current distribution at the electrode affects directly the uniformity of electroplated metals and the composition of
alloys and composites; it is therefore essential to be able to predict the current distribution at various length scales, starting from
the scale of the electrochemical cell (centimeters to meters) down to the mass transport scale (microns) and even to the spatial scale
of ET (nanometers).13
In first approximation, outside the diffusion layer the current distribution can be calculated assuming that the electrolyte concen-
tration is uniform, and the potential distribution can be calculated with the Laplace equation DF ¼ 0, where D is the Laplace oper-
ator, i.e., the sum of the second derivatives of the potential field F along the three spatial directions. Solution of this equation gives
the CD vector at each point in the electrolyte using Ohm’s law, which in one dimension reads
j ¼ sdF=dx (15)
where s is the electrolyte conductivity.
Three cases can be distinguished to determine the current distribution at various length scales:
(i) Primary current distribution: the electrode polarization at the electrode interface and mass transport are neglected; the CD can be
calculated by the Laplace equation and it depends only on cell geometry.
(ii) Secondary current distribution: the potential drop at the electrode due to the double layer cannot be neglected. The CD distri-
bution in this case is calculated from the primary distribution, corrected by the voltage drop at the electrode, which in turn is
determined via the local CD. This distribution is more uniform than the primary one because higher current densities lead to
higher voltage drops at the interface, decreasing the local CD. As such, the secondary distribution depends both on cell
geometry and electrode polarization h. The Wagner number Wa ¼ (dh/dj)/r$L represents the ratio between the polarization
dh/dj at the interface and the Ohmic drop in the bulk, r is the electrolyte resistivity, and L is a typical size of the electrode. If Wa
is large, the current distribution is close to uniform, while Wa ¼ 0 returns the primary current distribution. An electrochemical
system where the secondary current distribution is approximately uniform is said to have a high throwing power. The cor-
responding electrolytes ensure the formation of a uniform coating even at complex surface shapes.
(iii) Tertiary current distribution: under large applied overvoltages, the charge transfer rate may become faster than the ion transport
rate and the system is close to the diffusion limiting current; in these conditions, the current distribution depends both on the
potential distribution and the local rate of mass transport and is calculated accordingly.
Note that when more than one reaction is taking place in the same electrochemical cell, different reactions may be occurring under
different kinetic conditions, leading to different current distributions.
In summary, any inhomogeneity in the current distribution will result in thickness inhomogeneity, and in the case of alloys,
both in thickness and composition; it is therefore essential to predict these effects and, if needed, control the current distribution
by forced convection such as can be obtained with the RDE or with forced flow.
A variety of numerical methods have been developed to precisely calculate these current distributions; these algorithms have
been used for the prediction of thickness and composition distribution in different electrochemical cell geometries, distinct chem-
istries, and most importantly patterned substrates with applications in microelectronics and microsystems.
148 Fundamentals of Electrodeposition

Charge transfer mechanism and surface diffusion


When the hydrated metal ion approaches the compact double layer region ( 1–10 nm in thickness), the local electric field increases
strongly and reaches values of the order of 107–109 V/m, sufficient to orient the water molecules bound to the metal ion. Under
these conditions, the progressive stripping of water molecules and the interaction of the ion with the electric field result eventually
in a charge transfer process, which in principle could be modeled using the Marcus theory, modified to account for the electron
density at the electrode. From an energetic point of view, however, it is difficult to explain the occurrence and the observed rates
of charge transfer from the metal ion to the electrode, because the solvation energy of a metal ion ranges between 5 eV/ion (for
a monovalent ion such as Agþ) and 20 eV/ion (for a divalent ion, such as Cu2 þ), much higher than the average energy of the
ion at room temperature ( 0.1 eV).
Gileadi hypothesized that charge transfer could not occur by electron tunneling, which would result in the formation of a highly
unstable neutral atom in solution before it could be adsorbed, but could possibly do so instead by ion transfer.14 Recently, ab initio
modeling of Agþ electrodeposition at an Ag electrode showed that an energy minimum for the ion in the double layer region occurs
at 2.9 Å away from the electrode surface, sufficient to enable electronic interaction with the electrode. Furthermore, specific inter-
action between the Agþ 5s orbital and the sp band of the Ag electrode may be strong enough to explain the high rate of deposition
experimentally observed.15
The details of metal ion reduction have not yet been worked out in detail, but it is reasonable to assume that the metal ion may
gradually loose the hydration sheath and its charge, consecutively or simultaneously, adsorbing on the surface in a neutral or partly
reduced state, and may diffuse across the surface at a growth site.16 Discharge of the metal ion has been analyzed from the energetic
standpoint, considering the various types of surface sites, and it was concluded that the direct transfer to a terrace site should be
favored over deposition at a kink site. The energy landscape for surface diffusion is determined by the morphological features
and energy barriers at the surface, which will be discussed in the next subsection.
At low overpotential h, adatom diffusion is the RDS for film growth, while at intermediate h, the RDS is electron exchange. Under
the latter conditions, the driving force for deposition is given by the Tafel relationship, as discussed previously.

Incorporation in the Growing Film


Growth of a metallic film occurs at a substrate surface, which ideally would be atomically flat. In practice, real surfaces exhibit defects
due to entropic effects. A model often used to describe the substrate surface is the terrace–ledge–kink (TLK) model; the acronym
derives from the initials of the main defects present at such surface (Fig. 6).17
The density of defects at the surface increases with temperature, and above a critical temperature (the roughening temperature
TR), a phase transition occurs, called the roughening transition, where the step energy formation decreases significantly. In electro-
chemical environments, TR may be close to room temperature and the density of kinks can be quite high under conditions relevant
for practical electrodeposition processes.
The metal deposition process described by the reaction Meaqz þ þ ze / Mekink at a rough surface occurs by direct adatom attach-
ment at a kink site. At this site, the total energy at a kink site, because at this site, the total energy of the system remains unaltered and
the process can be repeated with the next atom, thus satisfying the requirement for a process nominally in equilibrium, which could
be described by thermodynamic quantities. Under these conditions, the most probable reaction path for the metal ion is the direct
discharge at a terrace, whereby it diffuses to the nearest ledge and attaches at a kink site. At low overvoltage, surface diffusion of the
adatom is the RDS, and an analysis of this process shows that the apparent transfer coefficient is half of that expected by the BV
equation. Direct attachment at the kink site instead is less probable, but this process follows the BV equation.
The metal deposition process at an ideally smooth surface, in contrast, requires the formation of a stable nucleus, which in turn
involves the overcoming of an energy barrier. In the continuum approximation, the free energy of formation of a hemispherical
nucleus with radius r at the substrate surface is in fact the sum of a bulk term DGb corresponding to the formation of the bulk solid,
DGb ¼ ð4=3Þpr 3 zF jhj (16)

Fig. 6 The TLK (terrace–ledge–kink) model for a simple cubic surface with a (100) orientation. Note the presence of terraces, ledges, kinks, ada-
tom, and vacancies. Copyright Giovanni Zangari.
Fundamentals of Electrodeposition 149

which varies with the applied overpotential, plus the additional surface energy,
DGs ¼ 4pr 2 g (17)
which takes into account the additional free energy of the surface atoms. The sum of the two terms exhibits a maximum DGc
corresponding to a critical radius (or critical number of atoms when considering the atomistic nature of the process) above which
the overall energy of the system decreases below that of the critical configuration, yielding a nucleus that could grow (Fig. 7). Note
that a higher applied overvoltage leads to a smaller critical nucleus, the size of which could be in principle as small as few atoms and
the shape of which would be essentially two dimensional (2-D).18
If the overvoltage necessary for nucleus formation is small, the subsequent growth could also occur at low h, under a kinetic
growth regime described by the j–V Tafel characteristics.19 The shape of the resulting nuclei at low h would be dictated by thermo-
dynamics, via the surface energy balance of the film vs. the substrate. Let us denote gm, gsm, and gs as the surface energy of the metal,
substrate/metal interface, and substrate, respectively. If gm þ gsm > gs, the film does not wet the substrate and the growth is 3-D
(Volmer–Weber growth); for the opposite condition instead, the film wets the substrate and the growth is 2-D even without
applying a high overvoltage.
The rate of nucleation R at such surface is related to the probability of formation of the critical nucleus:
R ¼ CexpðDGc =kB T Þ (18)
where C is a constant dependent on the number density of adsorption sites and the atomic sticking coefficient to the growing
nucleus.
Under a constant driving force (DGc), the rate of nucleation is assumed to obey a first-order kinetics
N ðt Þ ¼ N0 ½1  expðkt Þ (19)

where N(t) is the density of nuclei at time t, N0 is N(t / N), and k the nucleation rate constant. This expression is usually simpli-
fied to either an instantaneous nucleation behavior: N(t) ¼ N0, where k is large and all nuclei are generated at the same time, or
a progressive nucleation, where the nuclei are generated linearly over time: N(t) ¼ kN0t. The two nucleation modes lead necessarily
to distinct morphologies; the first results in nuclei with uniform size, while the latter will exhibit a distribution of nuclei sizes. These
nuclei are generally assumed to be generated at random locations, and to increase in size via surface diffusion and attachment of the
reduced adatoms to the edges of the existing nuclei. Each nucleus therefore over time generates a region of attraction, such that no
additional nucleation may occur around an actively growing island. Models have been developed to predict the functional form of
the current transient vs. time for various modes of growth (2-D vs. 3-D, instantaneous vs. progressive); the growth modes occurring
in a given experiment can therefore be monitored and investigated experimentally with the help of these models by simply moni-
toring the current vs. time transients, under the assumption of no parasitic reactions.20
If the energy barrier for nucleus formation is small, the film grows under kinetic conditions, and a layer by layer (step flow)
growth can be observed because the time for the formation of a new nucleus is significantly larger than the time for completion
of the monolayer. Budevski et al. for example performed experiments on Ag single crystals under these conditions to directly observe
the propagation rate of the monolayer and the nucleation rate.21

Fig. 7 (A) Schematics of the energy terms involved in the formation of a nucleus; the volume (bulk) energy decreases as r,3 while the surface
energy increases as r 2; the total energy exhibits a maximum at a critical radius above which any further addition of matter decreases the
overall energy further, stabilizing the nucleus. Taken from http://www.tf.uni-kiel.de/matwis/amat/semitech_en/kap_3/backbone/r3_3_3.html. (B) The
total energy of the nucleus is pictured as a function of the number of atoms making up such nucleus; note how an increase in overvoltage decreases
the energy barrier to be overcome to stabilize the nucleus. Taken from Schmickler, W. Interfacial Electrochemistry. Oxford University Press, 1996;
Chapter 10, p. 130.
150 Fundamentals of Electrodeposition

Film growth under kinetic conditions leads to a rate of growth dependent on the atomic density at the various growing facets:
assuming a uniform rate of atom attachment, denser planes (low-energy facets) will grow more slowly, and in contrast less-dense
plane will grow faster, thus growing out of existence. This behavior will result in films exhibiting low-energy facets to the electrolyte.
By increasing the overpotential, nucleation processes become more probable and may occur on top of an island before the
monolayer is completed. The atoms deposited on top of this island cannot jump down due to the presence of a so-called
Erlich–Schwöbel energy barrier generated at the edge of the island; this barrier therefore leads over time to an increase in film rough-
ness since the adatom cannot decrease film roughness by jumping down.
Most technologically relevant electroplating systems exhibit a high energy barrier for nucleation, requiring a large overvoltage
(> 100 mV) to start growth.19 Under these conditions, growth is fast and the role of adatom diffusion in growth is stifled, resulting
in increased attachment at the most available sites, and leading to outgrowth. At even higher overvoltages, metal reduction occurs at
a sufficiently high rate that ion diffusion in the electrolyte cannot refill the Nernst diffusion layer. Such diffusion-limited growth in
ion-depleted regions will result in preferred growth at local film protuberances, which may lead to dendritic formations.
Adatom attachment at random sites may also result in the generation of defects, the most probable being twinning, due to the
generally low energy of this type of defects with respect to others such as dislocations or interstitials, particularly for face-centered
cubic metals. The most common processes for the nucleation of new grains include the formation of twin boundaries or the coa-
lescence of incoherent nuclei. The resulting grains will compete for growth, eventually forming columnar grains separated by grain
boundaries (GBs) with lower density due to their lower crystallinity.
In order to achieve dense, smooth films by electrodeposition, it is essential to enhance nucleation density by using large over-
potentials, and to limit grain growth through the inhibition that could be provided by additives adsorbing at the surface, thus slow-
ing down the attachment of adsorbed atoms. Thermal motion-induced attachment/detachment of organic species at the film surface
could also limit film roughening by allowing atomic attachment and growth only when the organic species is desorbed.

Metal Deposition: Rationalization, Trends, and Some Examples

Electrodeposition of single metals can be understood starting from their chemical and electrochemical properties, including their
electronegativity, the redox potential, and the kinetic parameters for metal reduction (exchange current and Tafel slope). A rational
understanding of electrodeposition of metals throughout the periodic table has been provided for ED from aqueous solutions, but
it is still far from complete in other electrolytes, in particular ionic liquids.
Piontelli was the first to classify the electrochemical behavior of metals based on their solid state and (electro)chemical prop-
erties, using a Born–Haber cycle to distinguish the important elementary steps occurring during electrodeposition.22 In this theory,
later refined by Cavallotti,23 metals are classified as normal, intermediate, and inert, on the basis of an electrochemical electronega-
tivity RNI that combines the magnitude of the enthalpy of atomization, ionization, and hydration of the metal under consideration,
normalized by the same expression for the hydrogen reduction process. Normal metals exhibit a high exchange current for metal
deposition and a low exchange current for hydrogen evolution, while inert metals behave oppositely, and the intermediate ones
possess features in between. Therefore, a normal metal will deposit in general almost reversibly, while it will need a high overvoltage
to evolve hydrogen, and an inert metal will exhibit an opposite behavior. This classification is very useful as it provides a rationale
for the prediction of trends across the periodic table; the theory in particular is capable to explain the relative behavior of the depo-
sition of late TMs, as discussed later in this section.
The metals that can be plated from aqueous solutions are highlighted in a periodic table of the elements and shown in Fig. 8. In
order to be able to deposit a metal from an aqueous solution, its redox potential need to be not too negative with respect to the
redox potential of hydrogen; the evolution of hydrogen in fact will occur between 0 and –1 VSHE (in acidic conditions), depending
on the catalytic activity of the electrode, and this reaction will establish the electrode potential, making it difficult to achieve the
potential below which the material of interest can be plated. Typically, the most electropositive metal that can be plated from
aqueous solutions is Mn (E0 ¼  1.18 VSHE).
Many TMs can be electrodeposited, while other metals can be only codeposited with Fe group metals, as shown in Fig. 8. Most
metal ions that can be plated have a positive charge when dissolved in solution, while some others are stable as hydrated anions or
as anionic complexes; the latter can be deposited by anodic processes, due to their anionic charge. In the following, some details will
be discussed on the electrodeposition, properties, and applications of some of the most common materials being electroplated,
including each one of the three classes identified by Piontelli and Cavallotti.

Copper
Applications for electroplated copper include among others the formation of undercoating for plating on plastics, die casting, elec-
trorefining, manufacturing of PCBs, and decorative. In 1997, electroplated copper displaced evaporated Al as the material for inter-
connect lines within semiconductor chips, a technology that has enabled over time the ongoing miniaturization of electronic
devices, today down to minimum features of  20 nm.
Copper can be electroplated from solutions based on acidic sulfate with or without small amounts of Cl, alkaline pyrophos-
phate, or fluoborate. Acidic baths exhibit high current efficiency but limited ability to grow uniform films, i.e., they show low
throwing power. Electroplated copper for interconnects utilizes acidic copper sulfate with the addition of additives to enable
Fundamentals of Electrodeposition 151

Fig. 8 A periodic table of the elements that can be plated from aqueous solutions; the green and red lines circumscribe the metals that can be
deposited as pure elements. In contrast, the elements surrounded by broken blue lines can only be deposited in parallel with the deposition of an Fe
group element (Fe, Co, or Ni). Copyright Giovanni Zangari.

bottom-to-top growth in very small, extreme aspect ratio trenches without forming voids. Additives include polyethylene glycol
(PEG) as an inhibitor, and sulfur- or nitrogen-containing organic molecules to accelerate Cu reduction at regions of negative curva-
ture. Similar formulations are used in micro- and nanoelectronic applications. Cu is also extensively used for electroforming and
microsystem components.
Cu is an intermediate metal, generally providing a relatively high intrinsic deposition rate. Cu ions are stable in the Cu2 þ oxida-
tion state and the deposition mechanism involves reduction at the cathode by two successive single-electron exchange steps; the first
step is slow and therefore is the RDS. Chloride may stabilize Cuþ at the surface by forming an adsorbed layer of CuCl, resulting in
a different mechanism and faster deposition. Mechanical properties sought in Cu plates may vary from highly ductile to very high
strength; for this reason, different electrolytes as well as plating modes, including various types of pulse plating, have been devel-
oped to cover all possible engineering needs.
Various Cu alloys also find extensive applications for decorative and engineering purposes; among the most important are brass
(CueZn), bronze (CueSn), and CueNi. Brass (30 wt% Zn) is used for plating on steel for better adhesion of rubber, while bronzes
with 10–20% Sn are of golden color and are used in decorative finishes. Both of these alloys have been traditionally electroplated
from cyanide-based electrolytes, but due to its toxicity, more environmentally friendly formulations are being developed. CueNi
are mostly plated from pyrophosphate-based electrolytes.

Iron Group Metals (Nickel, Cobalt, Iron)


Nickel plating is among the most common electrodeposition processes and is used mostly as a part of a multilayered system
designed to increase wear and corrosion resistance, and also able to function as a diffusion barrier. Note that Ni is more noble
than low carbon steel (essentially iron) and cannot corrode, i.e., provide sacrificial protection to steel parts. This multilayer coating
is made of Ni, coated by a thin layer of Cr. Ni is also used as a structural material in electroforming, a near net shape electroplating
process that uses a shaped mold (or mandrel) to form the final part in a single piece. Electroplating is uniquely suitable to faithfully
reproduce the shape of the mandrel due to the diffusional nature of the metal ion deposition process. The most popular electrolyte
for Ni plating is the Watts bath, which contains Ni sulfate, chloride, and boric acid, as a buffering agent. Decorative bright Ni is
obtained by the addition of a variety of proprietary additives and brighteners, mostly complex carbon and/or sulfur compounds.
Cobalt has properties similar to nickel but it is much less used for the same purposes due to its much higher cost. Co is instead
used mostly in magnetic applications, since it exhibits a high saturation magnetization and a large magnetic anisotropy, which
makes Co and its alloys suitable for manufacturing permanent magnets.
Electroplating of Fe is less important than Ni and Co, due to the limited corrosion resistance, even if its cost is lower. However,
FeeNi alloys can be electroplated to form very soft magnetic materials with extensive uses in magnetic and electronic applications at
microscale, such as transformers, inductors, and electromagnets. Among TM alloys, NieCo alloys are also of interest for the poten-
tial increase in hardness and the potentially hard magnetic properties. The hardness and toughness of this material is utilized in the
formulation of improved electroforming processes. NiePd in addition is used as a high-performance contact with high conductivity
and hardness.
152 Fundamentals of Electrodeposition

Fe, Co, and Ni are inert metals, with an intrinsic low deposition rate. All exhibit a standard redox potential more negative than
that of hydrogen, resulting in electrodeposition of these metals occurring in parallel with hydrogen evolution. The accepted mech-
anism for electrodeposition of these metals in sulfate-based electrolytes involves three main steps:
Me2þ þ H2 O/MeðOHÞþ þ Hþ (20a)

MeðOHÞþ þ e/MeðOHÞads (20b)

MeðOHÞads þ Hþ þ e/Me þ H2 O (20c)

where the metal ion interacts strongly with water, to form a metal monohydroxide, which is eventually reduced and adsorbs at
the electrode; finally, a further reduction mediated by the hydrogen ion leads to metal reduction.

Zinc
Due to its low redox potential, Zn in form of a coating is used mainly as a sacrificial anode for low carbon steel surfaces. Zn is
a normal metal; as such it exhibits a high exchange current and therefore tends to deposit quickly and to form dendritic and spongy
films already at relatively small current densities; metal complexation or inhibition of growth is therefore needed to obtain smooth,
dense films. The most common process to plate Zn was until few years ago an alkaline chemistry, while currently it involves also
acidic chloride baths. The mechanism in both electrolytes is similar, with a Zn–X species adsorbing at the surface and acting as the
intermediate for Zn deposition. The species –X could be –OH for sulfate-based electrolytes, or eCl for chloride-based solutions.
Since Zn electrodeposition tends to form dendrites, in order to avoid this type of morphology large amounts of organic compounds
are added to the electrolyte. Earlier formulations used alkaline electrolytes containing a variety of additives to hinder deposition
such as polyvinyl alcohol, and quaternary nitrogen compounds to obtain shiny films. Pulsed current has been used to decrease grain
size and increase brightness. An important process involves continuous plating of Zn at high current densities to provide corrosion
resistance for car frames; the continuous reel-to-reel process utilized commercially simplifies manufacturing and decreases costs. The
electrolytes used in this process are acidic electrolyte containing chlorides or sulfates and require strong agitation. ZneNi ( 12%),
ZneCo, and ZneFe (15–30%) alloys have been extensively studied as coatings with enhanced corrosion resistance with respect
to Zn.

Alloy Deposition24

An alloy is formed by mixing together two or more metallic elements in varying fractions, with the technological purpose to enhance
the material properties or generate novel ones with respect to those of the single constituent elements. Such mixture of two elements
may exhibit one crystal structure (phase) or a blend of different phases within the volume of material. Electrodeposition of alloys is
practiced for the same reasons alloys are made in bulk material form: electrodeposited alloys usually enable large improvement in
mechanical properties, corrosion resistance, and magnetic response, among others.
In this section, the discussion will be limited to binary alloys; generalization to ternary or more complex alloys is mostly straight-
forward. Electrodeposition of an alloy M–N is carried out from an electrolyte containing suitable salts of M and N, a supporting
electrolyte to increase the conductivity, and possibly several complexing agents or additives. The electrolyte chemistry, the applied
potential, and temperature all play a role in the control of alloy composition and microstructure. The correlation between alloy
structure and properties vs. growth conditions is quite complex and not yet completely understood, due to the widely different
properties and electrochemical characteristics of distinct metals. The simplest approximation to predict alloy composition is to
assume that the two metal ion species Mz þ and Nw þ are deposited independently at the cathode, with each metal ion
obeying its own reduction kinetics.25 Assuming that the two metal ions are reduced at a rate given by the respective generalized Tafel
expression (10),
jM ¼ j0;M expðzaM f hÞ (21a)

jN ¼ j0;N expðwaN f hÞ (21b)

it is possible to determine the overall growth rate


 
dm=dt ðalloyÞ ¼ AM jM =z þ AN jN w (22)

where Ai is the atomic weight of element i, and the alloy composition


 
xM ¼ nM =ðnM þ nN Þ ¼ ðjM =zÞ= jM =z þ jN w (23)

where x is the molar fraction, and n stands for the number of moles. The above equations disregard the possibility of spurious
faradaic reactions to occur, such as the hydrogen evolution reaction (HER). If the applied potential is sufficiently negative to enable
HER, its partial current should be considered and a faradaic efficiency for metal deposition should be calculated.
Fundamentals of Electrodeposition 153

Various distinct patterns of behavior for the alloy composition vs. applied potential are predicted by assuming independent
metal deposition and varying the redox potential, the exchange current, the Tafel slope, as well as the metal ion concentration,
which dictates the diffusion limiting current. The predicted trends could be compared with the experimental data for alloy depo-
sition in order to gain a better understanding of the mechanism of electrodeposition of specific alloy systems. Fig. 9 shows some
examples of these trends, displaying for each case a set of reduction kinetics and the corresponding composition vs. applied
potential.
Fig. 9(A, B) shows the j–V curves for the metals A and B exhibiting different redox potentials, the same Tafel slope, without any
diffusion limitation. Note that for potential more positive than –E1, metal A is deposited exclusively (xA ¼ 1). At the potential –E1,
metal B starts to be codeposited and the alloy composition jumps to xA ¼ x1. The composition remains constant since the ratio of
the two partial currents as a function of potential is constant. Fig. 9(C, D) shows the case of two metals with different redox poten-
tial, different Tafel slope, and no diffusion limitation. For potentials between –E0 and –E1, metal A is deposited exclusively. Below
E1, metal B is codeposited and, due to the larger Tafel slope, the relative ratio of B in the alloy increases with more negative potential.
At the potential –E2, the two characteristics intersect, meaning that the two partial currents of A and B are the same; this does imply
that the fraction of A (and therefore of B) is 0.5 only in the case that the number of electrons necessary to reduce the two metals is the
same, as assumed in the figure. Fig. 9(E, F) describes two metals showing the same Tafel slope and different limiting currents.
Between –E0 and –E2, the deposit consists of pure A; below –E2, the partial current of B increases and the fraction of A progressively
decreases; at –E3, the partial currents are the same and the fraction of A is 0.5 (assuming same oxidation state of the A and B metal
ions); the A fraction continues to decrease below –E3 until B reaches its limiting current, thereafter showing a constant composition.

Fig. 9 Some examples of the relationship between the current–voltage (j–E) characteristics of single metals (left column) and the alloy composition
vs. applied potential (right column) under the approximation of independent codeposition (IC) of the single metals and 100% efficiency. Details are
provided in the text. Copyright Giovanni Zangari.
154 Fundamentals of Electrodeposition

Note that in this figure we have assumed a sharp transition between the Tafel behavior and the limiting current behavior; in reality
this transition is smooth and gradual.
Traditionally, the IC hypothesis has been referred to as “normal,” while deviations from IC have been traditionally classified in
three main classes: irregular, anomalous, and induced; despite its appellation, the “normal” class is by far the least common. In fact,
IC is the exception instead of the rule; this is owed to the various interactions occurring among the metal ions in solution, as well as
among the intermediate species and foreign molecules present in the electrolyte and/or adsorbed at the interface. Common reasons
for IC being invalid include the following: (i) the addition of a metal ion Nw þ to an electrolyte containing Mz þ varies the ionic
strength of the electrolyte and changes the electrochemical equilibria, especially in presence of complexing agents, possibly resulting
in an alteration of the free metal ions concentration and therefore of the limiting current; (ii) intermediate species of the two metal
ions formed during electrodeposition may present a different solubility and therefore would adsorb at the electrode to a different
extent, varying the degree of adsorption of the various species and therefore the relative deposition rate and the alloy composition
with respect to the IC case; (iii) in aqueous solutions, when the applied potential at the cathode is more negative than the disso-
ciation potential of water, hydrogen is being generated in parallel with the reduction of the two metal ions; this results in alkalin-
ization of the electrolyte close to the electrode, increased stirring, and transport due to H2 bubble formation, all of which could
contribute to changes in composition. Additionally, in some cases (most commonly Cr, Cd, Pd) hydrogen is incorporated in the
growing film, which may cause embrittlement of the deposit.
Interactions at the interface are especially problematic to understand and predict, due to the difficulty to detect often short-lived
chemical species in situ; at the same time, understanding these effects is essential to rationally develop electrodeposition processes.
A classic example of interactions at the interface is the so-called anomalous codeposition,26 which occurs during the codeposition of
late 3d TMs, including Fe, Co, and Ni. In section “Metal Deposition: Rationalization, Trends and Some Examples” it was discussed
that these metals electrodeposit via a mechanism involving intermediate adsorbed species such as Me(OH)ad (Eq. 20). It has been
shown that under equilibrium conditions, the potential at which the Me(OH) species adsorb is most positive for Fe and more nega-
tive for Co and then Ni, leading to a larger extent of coverage at the electrode by Fe(OH)ad. This results in a faster reduction rate for
Fe, and a slower one for Co or Ni. Note that this ranking of deposition rate is opposite than that predicted based on the standard
redox potential (EFe0 < ECo0
< ENi0), motivating the “anomalous” nature of this process. While this phenomenon results in a slower
deposition of Ni in presence of Fe2 þ, it has further been recognized that the reduction rate of Fe is also accelerated in presence
of Ni2 þ; this phenomenon is not yet clearly understood, but has been modeled by hypothesizing the presence of a mixed complex
of the two species, formed at the surface.27 Anomalous codeposition has also been demonstrated for the codeposition of Fe, Ni, or
Co with Zn, and it is probably active also for other first row TMs.
Another class of unexpected codeposition behavior that has been understood in terms of interactions at the interface is the
induced codeposition process.28,29 Metallic elements such as Mo or W cannot be deposited alone in aqueous solutions, but can do
so if codeposited in presence of a TM (Fe, Co, or Ni). Again, the mechanism of induced codeposition is not yet completely under-
stood, but a general picture of the process calls for alloy deposition via two parallel paths: the first is the reduction of the TM through
its own complex, while the second involves interaction at the electrode between the TM and the nonplateable metal, resulting in
formation of mixed metals complex, which enable codeposition. The maximum Mo or W fraction in the alloy is up to 50 at.%,
suggesting the formation of an equiatomic complex, hypothetically of the form (TM)(XO4)(Cit)H 2 , where X is Mo or W. In
a similar manner, also metalloid/nonmetallic elements such as P, B can be electro-codeposited with a TM in presence of a reducing
agent such as hypophosphite H2 PO2  or dimethylamine borane (CH3)2NH $ BH3, respectively. Codeposition of Ni and P has been
extensively studied, particularly in the context of the electroless process; various models have been proposed over time but a defin-
itive mechanism has not yet been agreed upon.30 A general mechanism, suitable as a starting point to develop a detailed mechanism
for a definite material combination, has been proposed; this includes adsorption of the reductant R–H at the electrode (R– being for
example HPO2  for the hypophosphite), followed by dissociation, forming a radical R$ads that may react with hydroxyls, thus
generating electrons that could carry out metal reduction. The possibility to form metallic P codeposited with Ni could thus be
related to the peculiar configuration and energetics of adsorbed species.
An additional, unique feature of alloy electrodeposition is that the process of atom incorporation in the alloy is also affected by
the atomic bonding strength in the solid state, as quantified by the free energy of mixing of the alloy, DGmix. If M and N in the bulk
form a stable solid solution M–N, the bonding strength in the alloy is larger than the bonding in the pure metallic elements M and
N. The stronger bond enables metal reduction (M or N) at a more positive potential than the redox potential for metal deposition
on itself. The redox potentials of either metal during alloy deposition is therefore shifted in the positive direction with respect to the
elemental redox potential by the quantity DE ¼  DGmix/zF, where DGmix and therefore DE are composition dependent. The
phenomenon is referred to as underpotential codeposition (UPCD).31 The UPCD shift is obtained by an expression derived from
the Nernst equation, corrected by the activity of the metal in the alloy. For the reduction of Mz þ and incorporation in the alloy
Mzþ þ ze /Malloy (24)

the corresponding redox potential


  
Eeq ðMÞalloy ¼ E0 Mzþ þ ðRT=zF Þln aMzþ aMalloy (25)

is more positive than that of the reaction Mz þ þ ze / M, due to the activity of the M in the alloy, aMalloy, being < 1. UPCD in
particular allows to shift the redox potential of the more reactive element in the positive direction, facilitating the codeposition with
Fundamentals of Electrodeposition 155

more noble metals. An additional feature is that this formula can also be used to predict the alloy composition as a function of the
applied potential, assuming that alloy formation occurs sufficiently slowly to approximate thermodynamic conditions. UPCD has
been thoroughly validated on a number of alloys (Pt, Pd, and Au based), utilizing bulk thermodynamic data to predict alloy compo-
sition vs. applied potential. In particular, the composition vs. potential observed in the electrodeposition of AueCu alloy from non-
complexing solutions precisely fitted the theoretical expression Eq. (25), reproducing closely the whole range of composition vs.
potential.

Metal Oxides and Chalcogenides

Metal oxides and chalcogenides are metal compounds with oxygen or sulfur, selenium, tellurium, in a well-defined stoichiometry.
These classes of materials take advantage from the large anion electronegativity to finely vary the electronic structure and therefore
tailor material properties for example from conductor, to semiconductor to insulator, finding usage among others in semiconductor
devices for electronic or optical applications, enabling magnetic functionalities, sensing and data storage, and catalysis.32–34
Synthesis of metal oxides or chalcogenides by electrodeposition presents distinct advantages with respect to bulk material
production, since it can be carried out at low temperature, with limited energy expenditure and low material loss. Furthermore,
precise control of material morphology is possible through the adjustment of deposition conditions and the chemistry of the
electrolyte.
The electrodeposition of metal compounds occurs via two main processes35: the first is an underpotential deposition process
similar to that discussed in the case of alloys, whereby the free energy of formation of the compound shifts the reduction potential
of the less noble metal in the positive direction by the quantity DE ¼  DG/zF, enabling ideally the formation of a pure
compound.36 Within the potential range between the deposition onset of the less noble element, ELN, and (ELN þ DE) both
elements can be deposited, and formation of the stoichiometric compound can be achieved upon adjustment of the metal ions
concentration. The first metal compound system synthesized by electrodeposition was CdeTe, a II–VI semiconductor. Due to
the limited solubility of Te, acidic electrolytes were used; a typical deposition mechanism for CdeTe involves the reactions
Cd2þ þ 2e /Cd (26a)

HTeO2 þ 4e þ 3Hþ /Te þ 2H2 O (26b)

Cd þ Te/CdTe (26c)
While the elements alone deposit at  0.4 VSHE and þ 0.55 VSHE, respectively, the enthalpy of mixing of the compound
(DG ¼  98.4 kJ/mol), shifts Cd deposition from  0.4 to þ 0.1 VSHE. Fig. 10 shows the schematics j–V traces for the single metals
and the compound, highlighting the shift of Cd deposition due to the enthalpy of mixing and the range of potential available for the
formation of the compound.

Fig. 10 Current density vs. voltage traces for the deposition of Cd (top), Te (middle), and CdeTe compound (bottom). Note the potential windows
where CdTe and CdTe þ Cd can be deposited.36 Taken from Lincot, D. Electrodeposition of Semiconductors. Thin Solid Films 2005, 487(1), 40–48.
156 Fundamentals of Electrodeposition

Fig. 11 A typical ECALE cycle for the deposition of ZnS.38 Top: cyclic voltammetry of Na2S in ammonia buffer solution at a Ag(111) single crystal
electrode; the oxidation of S 2  to S occurs at about 0.75 VSCE. Bottom: cyclic voltammetry of 0.5 mM ZnSO4 at a S-covered Ag(111) single crystal,
showing Zn 2 þ reduction to Zn at 0.59 VSCE. Deposition cycles are discussed in the text. Taken from Innocenti, M.; Pezzatini, G.; Forni, F.; Foresti,
M. L. CdS and ZnS Deposition on Ag(111) by Electrochemical Atomic Layer Epitaxy. J. Electrochem. Soc. 2001, 148(5), C357–C362.

Ideally, in the potential interval between the deposition of Te and the deposition of Cd alone, the Cd deposition rate will be
thermodynamically limited such as to form the stoichiometric compound; in practice, the composition may be slightly off-
stoichiometric due to the limitations related to the rate of arrival of the metal ions. In order to optimize the stoichiometry, it
has been shown that Cd ions need to be present in solution in significant excess with respect to Te.
The second synthetic route to metal compounds is electrochemical atomic layer epitaxy (ECALE),37 whereby the constituent
elements are deposited ideally monolayer by monolayer at underpotential, using suitable single metal ion electrolytes that are peri-
odically exchanged in a thin layer cell where the substrate is polarized at the underpotential of choice, selected to deposit a monoa-
tomic layer of the target element while avoiding the dissolution of the existing film. Experimentally, such procedure in most cases is
capable to deposit  10 ML epitaxially, but above 100 ML, the film starts to present significant defects and the properties suffer.
Fig. 11 shows a typical ECALE cycle suitable for the deposition of ZnS.38 The first step is sulfide oxidation to sulfur at underpoten-
tial: S2 / S þ 2e, leading to the formation of a sulfur monolayer on Ag(111); successively the sulfide solution is displaced by a rinse
solution, followed again by the Zn solution; the reductive UPD of Zn onto S-covered Ag(111) is now carried out down to  0.75 VSCE,
using a potential window suitable to avoid S reduction. Note that the oxidative deposition nature of S allows to stabilize the Zn at
cathodic potentials during the S deposition, and the same occurs during Zn deposition. The process could thus be repeated.

Electrodeposition of Oxides
Electrodeposition of metal oxides can be carried out by either anodic or cathodic paths; in the former case, the metal ions are
oxidized to an oxidation state where they tend to precipitate and form an insoluble deposit. An important example of this process
is the electrodeposition of MnO2 from Mn 2 þ in acidic sulfate solutions according to the following reaction:
Mn2þ þ 2H2 O/MnO2 þ 4Hþ þ 2e (27)
Cathodic electrodeposition in contrast may be achieved by one of two routes; the first consists in the partial reduction of the
metal ion M from the oxidation state þ n to the þ 2y (2y < n), reaching conditions that enable the reaction with water and therefore
oxide precipitation:
Mnþ þ ðn  2yÞe þ xH2 O/MOy þ 2yHþ (28)
Fundamentals of Electrodeposition 157

For instance, Cu2O is synthesized via the following reaction:


2Cu2þ þ 2e þ H2 O/Cu2 O þ 2Hþ (29)
2þ þ
whereby Cu is first reduced to Cu electrochemically, the latter reacting with water to form Cu2O. 39

A distinct route for cathodic metal oxide formation involves the generation of base via a number of reactions, most commonly
including the reduction of water (30) or the reduction of nitrate (31).40
2H2 O þ 2e /H2 þ 2OH (30)

NO  
3 þ H2 O þ 2e /NO2 þ 2OH

(31)

The increase in pH due to hydroxide formation finally induces the precipitation of the metal oxide. The formation of ZnO is an
instructive example of this procedure; growth in this case proceeds through one of the two reactions (30) and (31), followed by
a two-step precipitation process:
Zn2þ þ 2ðOHÞ /ZnðOHÞ2 ðhydroxideÞ (32a)

ZnðOHÞ2 /ZnO þ H2 O ðdehydration to oxideÞ (32b)

Due the reversibility of the reactions (32), ZnO easily forms large single crystals ( mm scale), often with hexagonal columnar
shape, that underline the high quality of the material that can be achieved by this method.
An important difference between the electrodeposition of metallic materials and that of metal compounds lies in the nucleation
process. As discussed earlier, metallic films are formed by distinct metal ion reduction processes, leading to atom by atom incor-
poration in the film. The electrodeposition of compounds in contrast occurs through the nucleation of colloidal particles, which
coagulate at the electrode to form a solid deposit. According to the DLVO theory,41 a repulsive Coulomb force and an attractive
Van der Waals interaction compete, eventually leading to flocculation of colloidal particle and film formation.42 It is not clear
yet under what conditions nucleation would occur homogeneously (in the electrolyte), heterogeneously (at the electrode) or via
a combination of the two.

Electrodeposition of Chalcogenides
Chalcogenides materials mostly exhibit semiconductive properties, and as such are suitable candidates for solar cells, optoelec-
tronics, sensors, and thermoelectrics. The formation of chalcogenides is in principle quite similar to the electrodeposition of oxides
and therefore enjoys the same advantages such as the low temperature and simplicity of the synthesis. In particular, the possibility of
composition modulation in situ, during growth by the periodical variation of the applied potential, allows in some cases direct
formation of p–n junctions. The large number of existing chalcogenides materials and the wide range of their properties hinders
a comprehensive coverage; in the following, we will focus on the most representative materials from a historical and current interest
perspective.
Cd based: CdeTe, discussed earlier, is a prototypical semiconductor material, used for instance in infrared applications, X-ray and
gamma detectors. As discussed earlier, it is usually deposited by a one-step process at high temperature ( 80 C) from an acidic
solution containing CdSO4 in molar quantities and HTe2O þ in millimolar quantities; however, in order to obtain high quality,
thin epitaxial films ECALE has been employed. Under the conditions of the one-step process n-type doping is usually observed,
due to excess Cd positioned in interstitial sites, while for more positive potentials the material is p-type due to Cd vacancies.
Due to a larger ohmic drop across an increasing film thickness, it is necessary to shift the applied potential over time, which often
leads to the formation of an n–p junction in the film, which in turn may change the solid state current–voltage response of the
film.43
Homologous materials such as CdSe and CdS can be deposited from similar acidic electrolytes, with some differences due to the
distinct chalcogenide properties. For example, CdSe deposition suffers from excess Se deposition, which could be avoided by adding

SeSO3 2 .44 With regard to CdS, acidic solutions containing Cd2 þ and S2O 3  enable deposition by decomposition of the latter
compound to form colloidal sulfur. Alternatively, CdS can be advantageously grown by using ECALE, where a clear layer-by-layer
growth by in situ STM has been observed.45 The above compounds exhibit different band gaps Eg, decreasing with increasing atomic
number of the chalcogenide (CdS: 2.42 eV, CdSe: 1.84 eV, CdTe: 1.61 eV), and therefore they are used for different applications:
CdSe is used as an absorber layer in quantum dot solar cells, while CdS is used as a n-type window layer to be coupled with
a p-type absorber.
Cu(InGa)Se2 and Cu2ZnSnSe4: In recent years there has been a renewed interest in solar cell technology using advanced polycrys-
talline thin film absorbers (few micrometers thickness) instead of Si single crystal wafers (200 mm thick), characterized by less mate-
rial usage and lower cost with respect to the traditional technology. These materials include chalcopyrite Cu(In,Ga)(Se,S)2 (CIGS)46
and kesterite Cu2ZnSn(Se,S)4 (CZTS)47,48 semiconductor compounds including 3–5 elements (the elements placed in parentheses
can be exchanged in any fraction to tune properties).
CuInxGa1  x(Se,S)2 exhibit a chalcopyrite structure and band gap increasing with Ga fraction, varying between 1 and 1.7 eV for
pure selenide and 1.2–2.38 eV for pure sulfide, with In and Ga substituting each other. The electrodeposition of these compounds is
made difficult by the wide difference in redox potentials of the elements (Cu 0.34 VSHE, In  0.34 VSHE, Ga  0.53 VSHE,
158 Fundamentals of Electrodeposition

Se þ 0.75 VSHE) and their intrinsic reduction kinetics, being slowest for Se due to the four-electron exchange required. Deposition is
performed at low pH to enhance solubility and uses simple salts of the metal ions, or in some cases complexing agents such as citrate
or pyrophosphate. Various processes alternative to the direct codeposition of all elements have been demonstrated since the direct
process is difficult to accomplish. The deposition of oxides of the metallic elements followed by sulfurization/selenization is a little
bit easier due to the more positive potential ranges necessary to form oxides, but the control of composition remains difficult. Better
composition control and improved uniformity are achieved by the successive electrodeposition of stacked single metal layers,
binary or ternary alloys of the metallic element, or stacked deposition of binary sulfides or selenides. Electrodeposition processes
have been implemented demonstrating up to 21% efficiency on laboratory samples, and around 16% for larger areas.
These materials, however, suffer from several drawbacks; they contain potentially toxic materials and their abundance in the
earth’s crust could be insufficient to sustain a major technology development at the global scale; for this reason alternative semi-
conductor materials have been studied, among which the CZTS appears the most promising. The constituent elements being
used are for the most part nontoxic and are produced intensively already for other applications; electrodeposition processes similar
to those discussed for CIGS have been proposed and implemented. A major difference with respect to CIGS is that kesterite CZTS
materials exhibit a much more limited window of stability of the target phase, and therefore very precise conditions must be used to
obtain the structure sought after. Even if the pure target structure were achieved, over time it could dissociate into various binary and
ternary materials, with an overall lower free energy. Research in this respect is still ongoing.

Evolution of Film Structure and the Control over Morphology


The morphology of oxide and chalcogenide films is determined at the nucleation stage by the particle size and shape of the aggre-
gating particles and later by the growth conditions after nucleation. Predictive relationships are not available, but some guidelines
are available. The particle size and shape can be qualitatively controlled by the electrolyte chemistry, in particular pH and ionic
strength; as an example, small particles are favored by precipitation at a pH far from the point of zero charge in solutions of
high ionic strength.49 Control of film morphology can be achieved by use of suitable additives. Just like in electrodeposition,
the evolution of film morphology at low overpotentials, i.e., close to thermodynamic conditions, is determined by the relative
surface energies of the various crystallographic facets.47 Since the crystal seeks to minimize surface energy of the exposed surfaces,
high energy surfaces will tend to grow faster and be eliminated as growth proceeds, while low energy surfaces would grow larger in
area. Specific additives adsorbing preferentially to certain facets of the growing crystals may lower their surface energy, providing
a route to alter the preferential direction of growth. Choi50 for example demonstrated that Cu2O tends to form cubic crystals in
the absence of any additives while the adsorption of sodium dodecyl sulfate stabilizes the {111} facets and promotes the growth
of octahedral-shaped crystals. At high overpotentials instead, the diffusion of reactive species to the growth sites is the rate-
determining process, and a branching growth is then preferred. The two growth modes can be used in conjunction to finely control
the shape of oxide crystals. Similar processes are responsible for the wide variety of structures, from nanorods to nanoplates,
observed in ZnO formed by the addition in solution of KCl, NH4F, NH4CH3COO, and ethylenediamine (EDA).51

Electrodeposition of Composite Materials

Since the 1970s, electrodeposition from electrolytes containing metallic salts together with suspended micro-/nanosized particles
has been practiced, leading to particle incorporation in the electrochemically grown solid film and thus forming a composite. This is
a particularly simple process to incorporate particles in metallic films with the purpose to achieve increased microhardness,
improved tribological properties, corrosion resistance, lubrication, or add new functionalities such as magnetic or catalytic
properties.52
A fundamental understanding of the incorporation process is essential to be able to control and predict the volume fraction and
location of the particles in the films. Experimental work showed that the volume fraction of codeposited particles depends on the
particle concentration in solution and the applied CD (or potential); modeling of this process was first discussed by Guglielmi.53 He
postulated a two-step adsorption process, the first a physical process where the particles are still hydrated, and the second a field-
assisted, stronger adsorption that stabilizes particles on the surface of the electroplated film that are eventually covered and incor-
porated by the growing film. The method is widely used due to its simplicity as it depends only on three parameters related to the
strength of adsorption and of the electrical (DLVO type) attraction. The model provides a quantitative prediction of the effect of
particle concentration and CD on the extent of particle incorporation.
Fransaer54 modeled instead the particle adsorption and incorporation process at a RDE, by considering the trajectory of non-
Brownian particles (1 mm and larger); a particle stays at the electrode if the ratio of the normal vs. tangential force is favorable,
and later adsorbs through a specific DLVO force (of range  10 nm), contrasted by repulsive hydration forces. Particles suspended
in an electrolyte acquire a positive or negative charge depending on their potential of zero charge; for pH below the pzc the particle is
positively charged, and above is negatively charged, determining thus whether the particle may be attracted or not to the electrode.
Particles being incorporated could vary in size from few nanometers to 1 mm, and may include dielectric materials (Al2O3, SiC,
Si3N4, diamond, ZrO2, TiO2), metals (Cr), and polymers (polystyrene). The fraction of incorporated particles varies with the nature,
size and shape of the particle, the electrolyte chemistry (pH, surfactants), CD, hydrodynamic conditions, and temperature. Pulsed
current leads in general to a higher nanoparticle concentration; in particular, reverse (anodic) pulses lead to higher particle
Fundamentals of Electrodeposition 159

concentration due to the partial dissolution of the film during the anodic step; addition of ultrasonic agitation may further increase
the particle loading. Smaller particles are incorporated to a larger extent. Particle codeposition on the other hand may disrupt
growth of the metallic matrix and for example decrease the grain size of the metal, facilitating nanostructuring (and possibly
mechanical hardening) of the matrix. Increasing agitation usually enhances particle codeposition, but simple electrolytes in
some case may be unable to incorporate significant particle fractions; in these cases, surfactants may increase adsorption strength
and thus the fraction of incorporated particles.
Main applications of composite coatings include hardening of electroplated coatings for hardfacing or tribological application,
self-lubricating coatings by incorporating low friction particles (MoS2, PbS) or oil-filled microcapsule, an increase in corrosion resis-
tance due to particle incorporation or changes in the microstructure of the matrix, and application in catalysis, where the oxides
could be used as a support and nanoparticles would perform catalytic processes. Finally, composites of a functional material in
a matrix would provide additional functionality; examples include semiconducting or oxide powders in polymers or metals, for
photocatalysis, photoelectrochromism, and also energy storage.55

Nanostructured Materials

The typical material dimensions below which properties become sensitive to size are roughly 100 nm. If at least one dimension is at
or below this size range, it is expected that the properties of the material may be very different from the corresponding bulk mate-
rial.56,57 The variation in properties may be originated either by a change in electronic structure or by the high percentage of atoms
present at a surface or interface; for example, it can be calculated that a material with 5 nm grain size has a  50% fraction of atoms
at a GB.
Potential changes in properties vary widely based on the type of materials. Among metals, nanocrystalline materials exhibit
enhanced strength and hardness; for example, the yield strength sy is well known to vary with grain size, following the Hall–Petch
(HeP) relationship58:
sy ¼ s0 þ kda (33)

where the exponent a is usually 0.5 but can vary from 0.3 to 0.7. The strengthening with decreasing grain size is related to how
dislocations move in smaller and smaller grains; this behavior, however, can be extrapolated only down to 10 nm; below that an
inverse HeP behavior due to GB sliding occurs, decreasing the strength. The latter effect has been controversial for a while; recently,
however, it has been clarified that an inverse HeP behavior occurs which seems to be due to GB sliding resulting in the emission of
dislocations from GB junctions.59
Mechanical properties are also affected by porosity p, with the Young’s modulus E (E0 for a dense film), yield stress and
ductility affected as E/E0 ¼ 1–1.9p.60 Ductility is also affected negatively by grain size, due to the presence of pores, tensile insta-
bility, and crack nucleation. Electrical resistivity r is also affected by grain size, with lower grain size increasing r due to the
increase in scattering events at GBs. Melting point also depends on crystal size, with melting occurring first at GBs due to the
weaker bonding. Properties such as corrosion resistance or saturation magnetization of ferromagnetic materials in contrast do
not change much with grain size. Coercivity, however, decreases strongly with grain size due to the anisotropy averaging among
strongly interacting grains below the exchange length Lex.61,62 Magnetic nanoparticles below this size become superparamagnetic.
Finally, in semiconductor nanocrystals the band gap increases with decreasing size due to the change in electronic structure with
increasing confinement. Metallic nanoparticles also exhibit an altered electronic structure, enhancing catalytic activity and/or
generating plasmonic activity.
Electrodeposition is an advantageous method for the synthesis of nanomaterials, including metals, metal oxides, and compos-
ites. Nanostructured materials by electrochemistry can be obtained under conditions favoring the nucleation over existing grains
instead of grain growth. This in turn can be favored by (i) accelerating the nucleation rate via an increase in overpotential; (ii) using
complexing agents to reduce metals at high overvoltages; and (iii) using surface active species to further inhibit growth by slowing
down mobility of adatoms. Metals with low exchange current may further exhibit small grain size due to the large overvoltage neces-
sary to achieve significant current densities. Alloys on the other hand can be nanostructured under less stringent conditions as the
renucleation is favored when the atomic volume differences of the two constituent elements is sufficiently large that the growth of
a nucleus is hindered due to the large internal stresses. In contrast with most other nanostructure synthesis methods (inert gas
condensation, mechanical alloying, crystallization from amorphous solids, severe plastic deformation) electrodeposition is capable
to minimize porosity and therefore the trends of properties discussed above may be different. Other methods to enhance nanocrys-
tallinity is to use pulse plating, whereby a negative pulse may be capable to partly dissolve the film at regions of high free energy
(high stresses, protrusions), thus tailoring the grain size and morphology. Nanostructuring of a metal is also favored by codeposi-
tion with colloidal particles (SiC, SiN, C, Al2O3, microcapsules).63

See also: Electrocatalyst Preparation by Electrodeposition; Electrodeposition of calcium phosphates, oxides, molecules on metallic biomaterials; Ionic
Liquids in the Field of Metal Electrodeposition.
160 Fundamentals of Electrodeposition

References

1. Hunt, L. B. Gold Bull. 1973, 6 (1), 16–27.


2. Raub, C. In Metal Plating and Patination: Cultural, Technical and Historical Developments; La-Niece, S., Ed.; Butterworth-Heinemann, 1993.
3. Davis, J. R., Ed. Surface Engineering for Corrosion and Wear Resistance; Vol. 751; ASM International, 2001.
4. Schlesinger, M., Paunovic, M., Eds. Modern Electroplating; vol. 55; John Wiley & Sons, 2011.
5. Datta, M.; Landolt, D. Electrochim. Acta 2000, 45 (15), 2535–2558.
6. Andricacos, P. C.; Uzoh, C.; Dukovic, J. O.; Horkans, J.; Deligianni, H. IBM J. Res. Dev. 1998, 42 (5), 567–574.
7. Romankiw, L. T. Electrochim. Acta 1997, 42 (20), 2985–3005.
8. Braun, T. M.; Schwartz, D. T. Electrochem. Soc. Interf. 2016, 25 (1), 69–73.
9. Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundamentals and Applications, 2nd ed.; Wiley, 2000.
10. Inzelt, G.; Lewenstam, A.; Scholz, F. Handbook of Reference Electrodes, Springer: Berlin, 2013.
11. Marcus, R. A. J. Electroanal. Chem. 2000, 483 (1), 2–6.
12. Devanathan, M. A. V.; Tilak, B. V. K. S. R. A. Chem. Rev. 1965, 65 (6), 635–684.
13. Dukovic, J. O. IBM J. Res. Dev. 1990, 34 (5), 693–705.
14. Gileadi, E. Chem. Phys. Lett. 2004, 393 (4), 421–424.
15. Pinto, L.; Spohr, E.; Quaino, P.; Santos, E.; Schmickler, W. Angew. Chem. Int. Ed. 2013, 52 (30), 7883–7885.
16. Despic, A. R.; Budevski, E. B. In Electrochemistry; Physical Chemistry: An Advanced Treatise, vol. IX B, Academic Press: New York/London, 1970; pp 611–729.
17. Ibach, H. Physics of Surfaces and Interfaces, Springer-Verlag: Berlin Heidelberg, 2006.
18. Budevski, E.; Staikov, G.; Lorenz, W. J. Electrochim. Acta 2000, 45 (15), 2559–2574.
19. Guo, L.; Oskam, G.; Radisic, A.; Hoffmann, P. M.; Searson, P. C. J. Phys. D: Appl. Phys. 2011, 44 (44), 443001.
20. Hyde, M. E.; Compton, R. G. J. Electroanal. Chem. 2003, 549, 1–12.
21. Budevski, E. B.; Staikov, G. T.; Lorenz, W. J. Electrochemical Phase Formation and Growth: An Introduction to the Initial Stages of Metal Deposition, John Wiley &
Sons, 2008.
22. Kozlov, V. M.; Peraldo Bicelli, L. J. Cryst. Growth 1999, 203 (1), 255–260.
23. Cavallotti, P. L.; Nobili, L.; Franz, S.; Vicenzo, A. Pure Appl. Chem. 2010, 83 (2), 281–294.
24. Brenner, A. Electrodeposition of Alloys: Principles and Practice, Elsevier, 2013.
25. Landolt, D. Plat. Surf. Finish. 2001, 88, 70–79.
26. Matlosz, M. J. Electrochem. Soc. 1993, 140 (8), 2272–2279.
27. Zech, N.; Podlaha, E. J.; Landolt, D. J. Electrochem. Soc. 1999, 146 (8), 2886–2891.
28. Podlaha, E. J.; Landolt, D. J. Electrochem. Soc. 1996, 143 (3), 885–892.
29. Eliaz, N.; Gileadi, E. In Modern Aspects of Electrochemistry; Author, C. D., Ed.; Springer: New York, 2008; pp 191–301.
30. O’Sullivan, E. J. Adv. Electrochem. Sci. Eng. 2002, 7, 225–274.
31. Brankovic, S. R.; Zangari, G. In Electrochemical Engineering across Scales: From Molecules to Processes, VCH-Wiley: Weinheim, Germany, 2015.
32. Yuan, C.; Wu, H. B.; Xie, Y.; David Lou, X. W. Angew. Chem. Int. Ed. 2014, 53 (6), 1488–1504.
33. Fierro, J. L. G., Ed. Metal Oxides: Chemistry and Applications, CRC Press: Boca Raton, FL, 2005.
34. Gao, M.-R.; Xu, Y.-F.; Jiang, J.; Shu-Hong, Y. Chem. Soc. Rev. 2013, 42 (7), 2986–3017.
35. Therese, G. Chem. Mater. 2000, 12 (5), 1195–1204.
36. Lincot, D. Thin Solid Films 2005, 487 (1), 40–48.
37. Stickney, J. L. Adv. Electrochem. Sci. Eng. 2002, 7, 1–106.
38. Innocenti, M.; Pezzatini, G.; Forni, F.; Foresti, M. L. J. Electrochem. Soc. 2001, 148 (5), C357–C362.
39. De Jongh, P. E.; Vanmaekelbergh, D.; Kelly, J. J. Chem. Mater. 1999, 11 (12), 3512–3517.
40. Peulon, S.; Lincot, D. J. Electrochem. Soc. 1998, 145 (3), 864–874.
41. Adamczyk, Z.; Weronski, P. Adv. Colloid Interface Sci. 1999, 83 (1), 137–226.
42. Holdich, R. In Fundamentals of Particle Technology, Midland Information Technology and Publishing: Loughborough, UK, 2002; pp 131–140.
43. Kampmann, A.; Cowache, P.; Vedel, J.; Lincot, D. J. Electroanal. Chem. 1995, 387 (1), 53–64.
44. Mishra, K. K.; Rajeshwar, K. Journal of electroanalytical chemistry and interfacial electrochemistry 1989, 273 (182), 169.
45. Demir, U.; Shannon, C. Langmuir 1994, 10 (8), 2794–2799.
46. Reinhard, P.; Chirila, A.; Blösch, P.; Pianezzi, F.; Nishiwaki, S.; Buecheler, S.; Tiwari, A. N. IEEE J. Photovoltaics 2013, 3 (1), 572–580.
47. Delbos, S. EPJ Photovoltaics 2012, 3, 35004.
48. Kumar, M.; Dubey, A.; Adhikari, N.; Venkatesan, S.; Qiao, Q. Energy Environ. Sci. 2015, 8 (11), 3134–3159.
49. Jolivet, J.-P.; Cassaignon, S. ; Chanéac, C. ; Chiche, D. ; Tronc, E. J. Sol-Gel Sci. Technol. 2008, 46 (3), 299–305.
50. Choi, K.-S. J. Phys. Chem. Lett. 2010, 1 (15), 2244–2250.
51. Xu, L.; Guo, Y.; Liao, Q.; Zhang, J.; Dongsheng, X. The Journal of Physical Chemistry B 2005, 109 (28), 13519–13522.
52. Roos, J. R.; Celis, J.-P.; Fransaer, J.; Buelens, C. JOM 1990, 42 (11), 60–63.
53. Guglielmi, N. J. Electrochem. Soc. 1972, 119 (8), 1009–1012.
54. Fransaer, J.; Celis, J.-P.; Roos, J. R. J. Electrochem. Soc. 1992, 139 (2), 413–425.
55. Boccaccini, A. R.; Zhitomirsky, I. Current Opinion in Solid State and Materials Science 2002, 6 (3), 251–260.
56. Arzt, E. Acta Mater. 1998, 46 (16), 5611–5626.
57. El-Sayed, M. A. Acc. Chem. Res. 2004, 37 (5), 326–333.
58. Hansen, N. Scr. Mater. 2004, 51 (8), 801–806.
59. Quek, S. S.; Chooi, Z. H.; Wu, Z.; Zhang, Y. W.; Srolovitz, D. J. J. Mech. Phys. Solids 2016, 88, 252–266.
60. Meyers, M. A. Prog. Mater. Sci. 2006, 51 (4), 427–556.
61. Erb, U. Nanostruct. Mater. 1995, 6 (5), 533–538.
62. Herzer, G. IEEE Trans. Magn. 1990, 26 (5), 1397–1402.
63. Robertson, A.; Erb, U.; Palumbo, G. Nanostruct. Mater. 1999, 12 (5), 1035–1040.

You might also like