120037171_E-ECHP_00_00_R2_081905

1
2
Electrodeposition
3 E
4
André Avelino Pasa
5
Thin Films and Surfaces Group, Departamento de Fı´sica, Universidade Federal de Santa
6 Catarina, Florianópolis, Santa Catarina, Brazil
7
8 Maximiliano Luis Munford
9 Group of Organic Optoelectronic Devices, Departamento de Fı´sica, Universidade Federal do
10 Paraná, Curitiba, Paraná, Brazil
11
12
13
14
15 INTRODUCTION interface process mediated by the occurrence of elec-
16 trochemical reactions that promote the reduction or
17 Most important concepts and techniques of an electro- the oxidation (redox reactions) of the ionic species.
18 deposition are introduced and described here in an An electrochemical cell with a battery is illustrated in
19 easy-to-understand way. Examples of technological Fig. 1, where the motion of the ions toward the
20 applications are given, with emphasis on the fabri- electrodes is also sketched. In this case, the metallic salt
21 cation of nanostructures. NiSO4 (nickel sulfate) dissolved in water is a practical
22 example of an electrolyte for Ni-plating metallic objects.
23 In this example, the object to be plated is a key, placed
24 DEFINITIONS AND HISTORY as the working electrode. By applying an external
25 voltage with the negative terminal of the battery
26 Electrodeposition is an electrochemical process that connected to the working electrode, the Ni2þ move to
27 allows the preparation of solid deposits on the surface this electrode, where deposition takes place, and the
28 of conductive materials. It is a commercially highly SO42
toward the positively charged counter-electrode.
29 relevant process, providing the basis for many indus- An essential characteristic of electrochemical reac-
30 trial applications, such as electro-winning, refining, tions is that the exchange of charge does not occur
31 and metal plating. Metal plating is the process that between chemical species, as it does in a typical chemical
32 has perhaps the closest contact with most people’s reaction, but between chemical species and the electrode.
33 everyday life, because we are surrounded by things that The electrochemical reaction that is most important for
34 have a protective or decorative coating, such as the electrodeposition process is the one that occurs at
35 watches, buttons, belt buckles, doorknobs, handlebars, the working electrode; i.e., for the example, in Fig. 1,
36 etc. Additionally and more recently, as will be seen it is the reduction reaction Ni2þ þ 2e
! Ni0 , where
37 below, not only do the circuit boards and the packaging the Ni ions are reduced by receiving two electrons (e
)
38 modules of computers, but also the recording and from the electrode. At the counter-electrode, the oxida-
39 reading heads of their hard disk drives and the micro- tion of the sulfate radical is too energetic to occur,
40 processor chip itself may have plated material on them. and the most probable oxidation reaction for inert
41 Electrodeposits are formed by the action of an elec- electrodes in an aqueous electrolyte is the electrolysis
42 tric current passing in an electrochemical cell, a device of the water, forming Hþ and O2 through the reaction
43 that consists of two conductive or semi-conducting H2 O ! 2Hþ þ 1=2O2 þ 2e
. This reaction occurs
44 electrodes immersed in an electrolyte. The electrodes by donating two electrons to the anode, completing
45 are called the working electrode (cathode), consisting the electrical circuit and keeping the electric charge
46 of the object where electrodeposition is planned, and balanced. Frequently used inert electrodes include
47 the counter-electrode (anode), necessary to complete platinum and glassy carbon. If the counter-electrode
48 the electrical circuit. Electrolytes for electrodeposition is a metallic bar or foil (a non-inert electrode), the
49 are usually aqueous solutions containing positive and electrodissolution of the metal could provide electrons
50 negative ions, prepared by dissolving metal salts. The for the electrode and ions for the solution.
51 electric current that flows between the two conductive Historically, the discovery of electrodeposition is
52 electrodes in the presence of an external voltage is attributed to Luigi V. Brugnatelli, an Italian professor,
53 because of the motion of charged species, via migration who in 1805 was able to electrodeposit gold on the
54 and diffusion, towards the surfaces of the polarized elec- surface of a metallic object, from a solution containing
55 trodes. At the surface of the electrodes, the conduction dissolved gold, using a voltaic pile (battery). About 40
56 mechanism must change from ionic to electronic, an years later, John Wright, from Birmingham, England,
Encyclopedia of Chemical Processing DOI: 10.1081/E-ECHP-120037171
Copyright # 2006 by Taylor & Francis. All rights reserved. 821
 120037171_E-ECHP_00_00_R2_081905
822 Electrodeposition

1 BATTERY This approach is relatively simple and inexpensive,

2 and is known as galvanostatic plating system, because
3 electrons – + electrons the current between the electrodes is controlled (main-
4 tained constant). Another important deposition mode
5 Working Counter is the pontentiostatic one. This mode is a consequence
6 Electrode Electrode of the development of electrochemical science, where
7 electrochemical reactions at the surfaces of electrodes
8 are carefully investigated. The electrochemist devel-
9 – oped reference electrodes, in order to measure the
Ni++ SO4
10 AQ1/ potential drop near the surface of electrodes. Assuming
11 PR that the electrolyte contains sufficient ions (has suffi-
Electrolyte
12 ciently high conductivity) to avoid any ohmic voltage
13 drop, because of the resistance of the electrolyte
14 ELECTROCHEMICAL CELL between the electrodes, all the voltage applied by the
15 battery (in Fig. 1) will appear near the surface of the
16 Fig. 1 Positively charged nickel ions in the electrolyte are electrodes, where a charged region is formed (usually
17 attracted by the negatively charged key (working electrode). named the double layer). It is very important to mea-
At the surface of the key they are reduced by gaining two
18 sure the voltage drop across these charged regions,
electrons, and metal is deposited.
19 because it controls the driving force for phase transfor-
20 mation from ion to reduced state. A simple metal foil
21 discovered that potassium cyanide was a suitable could be used as a reference electrode; however,
22 electrolyte for gold and silver electroplating. This because of the need to have a standard electrode to
23 discovery made electrodeposition an important com- measure potential drops at the surface of different
24 mercial process for covering the surface of various types of working electrode in contact with different
25 kinds of metallic object with thin coatings of metals electrolytes, a hydrogen electrode was elected, and
26 for corrosion protection and decorative purposes. Sub- now all electrode potentials are quoted relative to this
27 sequently, baths for the deposition of other metals and arbitrarily chosen reference electrode. What is always
28 alloys such as nickel, zinc, tin, and brass (an alloy measured is the potential difference between two
29 consisting essentially of copper and zinc in variable electrodes. By defining the potential of the hydrogen
30 proportions) were developed. For the next 100 years, electrode as zero, it is possible to generate a table of
31 the main idea was to use electrodeposition for covering all the possible redox reaction potentials relative to this
32 the surface of inexpensive materials with a thin layer of electrode, and these potentials are called standard
33 a noble metal. By the 1940s, however, electrodeposi- potentials. The standard hydrogen electrode (SHE),
34 tion was rediscovered by the electronics industry. The which is usually constructed by bubbling hydrogen
35 electrodeposition of gold for electronic components gas over an immersed platinum foil, has its operation
36 was a totally different kind of application of electrode- based on the redox reaction H2 $ 2Hþ þ 2e
.
37 position techniques. Other reference electrodes that are robust, stable, and
38 Over the years, electrodeposition became a highly easily constructed than the SHE are frequently used
39 developed process. Direct current (DC) power supplies in the laboratory, give potential measurements that
40 were developed; anodic and cathodic reactions were can be converted to standard potentials by adding or
41 described; new safer baths based on acid electrolytes, subtracting a constant value. The most common are
42 avoiding the earlier poisonous cyanide-based ones were the calomel electrode (Hg=Hg2Cl2) and silver=silver
43 discovered; models for the deposition process incorpor- chloride electrode (Ag=AgCl).
44 ating mass transport to the electrodes, charge transfer The potential of an electrochemical cell, also known
45 kinetics, and nucleation and growth at the working elec- as the cell potential or electromotive force (emf) is the
46 trode were developed; and regulatory rules for waste sum of the potential drops at the cathode and anode,
47 water emission and waste disposal were created. Simul- where the reduction and oxidation reactions occur.
48 taneously, a gradual improvement in electrodeposition With the introduction of a reference electrode the
49 for large scale manufacturing processes took place. potentials of these two electrodes can be measured
50 independently, allowing the independent investigation
51 of the reactions that are taking place at each electrode
52 ELECTRODEPOSITION APPARATUS (working or counter). These redox reactions are called
53 AND CONCEPTS half-cell reactions or simply half-reactions. The half-
54 reaction potential E 0 can be measured with a SHE
55 Electrodeposition on the industrial scale requires an electrode at standard conditions, i.e., at electrolyte
56 electrochemical cell and a DC current power supply. concentrations of 1 M, gas pressures of 1 atm., and
 120037171_E-ECHP_00_00_R2_081905
Electrodeposition 823

1 Table 1 Standard electrode potentials in aqueous solution at 25 C

2 Cathode half-reaction E0 (V)
3 þ
Na ðaqÞ þ e
! Na ðsÞ
2.71
E
4
5 2H2 O ðlÞ þ 2e
! H2 ðgÞ þ 2OH
ðaqÞ
0.83
6 Fe2þ ðaqÞ þ 2e
! Fe ðsÞ
0.41
2þ

7 Ni ðaqÞ þ 2e ! Ni ðsÞ
0.23
8 2Hþ ðaqÞ þ 2e
! H2 ðgÞ 0.00
9 2Cu 2þ
ðaqÞ þ 2e

þ 2OH

! Cu2 O ðsÞ þ H2 O ðlÞ 0.17
10
Cu2þ ðaqÞ þ 2e
! Cu ðsÞ 0.34
11 þ

12 O2 ðgÞ þ 4H ðaqÞ þ 4e ! 2H2 O ðlÞ 1.23
13 S2 O2

8 ðaqÞ þ 2e


! 2SO2

4 ðaqÞ 2.01
14 Note: aq., g and I denote aqueous, gas and liquid respectively.
15
16
17 temperature of 25 C, and tabulated. Table 1 shows a the experiment or deposition process, as depicted in
18 set of standard potentials for cathode half-reactions. Fig. 3A. An additional mode called pulsed deposition
19 The introduction of the reference electrode led to a is also illustrated in Fig. 3B. In this mode, for pulsed
20 different experimental setup for electrochemical and potential, the potentiostat switches the working
21 electrodeposition experiments. Fig. 2(A) shows an electrode potential between two values in order to have
22 electrochemical cell with three electrodes (working-, the potential varying as a square wave. For pulsed
23 reference-, and counter-electrodes) and a potentiostat. current deposition, a current source with a square wave
24 The potentiostat is an electronic apparatus that main- output is sufficient.
25 tains the potential difference between the working- and The three-electrode cell and potentiostat is also a
26 reference electrodes by controlling the potential differ- powerful experimental tool for electrochemical inves-
27 ence between the working- and counter-electrodes. tigations, permitting the implementation of different
28 Fig. 2(B) shows a block diagram of the electronic cir- techniques, such as voltammetry. This technique
29 cuitry of a potentiostat with an operational amplifier consists of applying a potential ramp to the working
30 that keeps the voltage between reference electrode electrode, which is achieved by applying a potential
31 (RE) and working electrode (W) equal to the applied ramp to the positive terminal of the operational
32 voltage E at the positive terminal, by regulating the cell amplifier [of Fig. 2(B)], and measuring the resultant
33 potential between W and counter-electrode (CE). By cell current. When the applied potential starts at a
34 convention W is connected to ground. defined level and comes back to the same value after
35 The three-electrode cell and the potentiostat enable a period of time, the technique is called cyclic volta-
36 the potentiostatic mode of deposition mentioned mmetry. When the applied potential starts at a level 1
37 above. The potentiostatic mode means that the poten- and goes to a level 2, the resulting plot of the current
38 tial of the working electrode is kept constant during versus the potential, is called a polarization curve or
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55 Fig. 2 (A) Electrochemical cell with three electrodes connected to a potentiostat. (B) Electronic sketch illustrating the mode of
56 operation of a typical potentiostat.
 120037171_E-ECHP_00_00_R2_081905
824 Electrodeposition

1 A B
2 POTENTIOSTATIC PULSED MODE
3 MODE (Potential / Current)
4 Eb / Ib
Potential

5
Applied

E
6 Ea / Ia Fig. 3 Illustration of the potential as a AQ1=PR
7 function of time for potentiostatic and
8 Time Time pulsed deposition modes.
9
10
11 simply voltammogram. Fig. 4(A) illustrates a poten- deposit that is very homogeneous, apart from a micro-
12 tial ramp applied to the working electrode, and scopic defect because of a hydrogen bubble, obtained
13 Fig. 4(B) the corresponding plot for the variation of at a deposition potential of
1.1 V. Fig. 5(C) shows a
14 the cathodic current as a function of the cathodic plot of the deposition current. This plot, known as a
15 potential. This figure shows a typical polarization current transient, gives information about the deposi-
16 curve for the reduction of a metal at a conductive tion process and allows the calculation of the elec-
17 electrode. The onset of current, indicated by an trodeposited charge from the area below the curve.
18 arrow, corresponds to the minimum cathodic poten-
19 tial at which the reaction occurs (a fundamental value
20 for electrodeposition purposes), the peak corresponds
21 to the maximum current at a given rate of change of ELECTRODEPOSITION MECHANISMS
22 potential (also known as the reduction peak), while
23 the third characteristic feature of the plot is the Electrodeposited Charge
24 saturation of the current at more negative potentials.
25 The intensity of the reduction peak depends on the Because the electrodeposition process involves the
26 potential scan-rate, this peak being not observed, in transfer of electrons to an electrode, by measuring
27 many cases, because of other reactions that occur the current in the electrochemical cell, it is in principle
28 simultaneously. possible to calculate the amount of material deposited.
29 If no other reaction occurs in parallel, then we can
30 assume that the reaction at the working electrode in
31 aqueous electrolyte is just the simple reduction of a
32 Potentiostatic Deposition
metal (M)
33
34 This section will describe the pontentiostatic mode
35 using an electrolyte containing cobalt sulfate. By Mnþ þ ne
! M0 ; ð1Þ
36 applying a potential ramp it is possible to obtain the
37 polarization curve shown in Fig. 5(A). In this figure where a metal ion Mnþ is reduced to a metal atom M0
38 the onset of Co reduction is about
0.8 V. By selecting after gaining n electrons. By assuming that all the
39 a deposition potential negative than this value, it is metal ions reduced at the surface of the working
40 possible to obtain a deposit. Fig. 5(B) illustrates a Co electrode stick to this surface, the total amount of
41 electrodeposited material can easily be calculated from
42 the charge Q (in coulombs), which represents the
43 product of the total amount of electrodeposited atoms
44 A B N times the charge of n electrons, as given by the
45 E I expression
Time Ef Ei
46 E
47 Ei Q ¼ Nne; ð2Þ
48
49
50 where e is the charge of one electron, equal to
51
1:6  10
19 C. The charge Q is calculated from the
AQ1/ current transient. If the deposition current is constant,
52 Ef
PR Q can be calculated by simply multiplying the current I
53
54 by the deposition time t,
Fig. 4 Illustration of the applied potential ramp (A) to
55 obtain a typical polarization curve (voltammogram) of metal
56 deposition on a metal working electrode. Q ¼ It: ð3Þ
 120037171_E-ECHP_00_00_R2_081905
Electrodeposition 825

1
2
3 E
4
5
6
7
8
9
10
11
12 Fig. 5 (A) Polarization curve of an electrolyte containing cobalt sulfate, (B) scanning electron microscope (SEM) image of a Co AQ1=PR
13 deposit obtained at
1.1 V, and (C) the corresponding deposition current as a function of time (current transient). The reference
14 electrode was saturated calomel and the working electrode semiconducting silicon. (From Ref.[1].)
15
16
17
However, if the current is varying during the deposi- This calculation gives a deposit thickness in meters
18
tion, as shown in Fig. 5(C), Q can be calculated by inte- that has to be converted into more adequate units
19
grating the current I ¼ IðtÞ as a function of time, as microns (1 mm ¼ 10
6 m) or nanometers (1 nm ¼
20
10
9 m). For a precise calculation of the thickness of
21 Z deposits, it is necessary to take into account possible
22 Q ¼ IðtÞdt: ð4Þ electrode reactions that occur simultaneously with
23
the main reaction. One very common example is the
24
hydrogen evolution reaction, 2Hþ þ 2e
! H2 .
25 In order to calculate the thickness h (m) of the deposit This reaction is so rapid, in conditions such as rela-
26 on a known area A (m2) of the surface of the electrode tively high cathodic potentials in acidic electrolytes,
27 where deposition occurred, the quantity N can be that it dominates the exchange of electrons at the
28 expressed as surface of the electrode. The contribution of hydrogen
29
evolution to the cathodic current must be determined
30 mNa
N ¼ ; ð5Þ in order to obtain the efficiency of the plating process.
31 M This may be done indirectly by measuring the thickness
32
of the deposit and calculating the amount of charge
33 where m is the electrodeposited mass (g), Na is corresponding to ‘‘missing’’ metal. The presence of
34 Avogadro’s number (the number of atoms in a mole, hydrogen during the deposition has several effects on
35 equal to 6:02  1023 ) and M is the atomic weight. the metallurgical properties of the deposits. One of
36 Using the density d (g=m3), given by d ¼ m=V, these effects is the formation of gas bubbles that mask
37 where the volume V (m3) is given by the product of the surface of the electrode locally, introducing
38 the area, and the thickness, V ¼ Ah, Eq. (5) can be microscopic defects in the electrodeposited layers [see
39 rewritten as example of this effect in Fig. 5(B)].
40
41
dAhNa
42 N ¼ ; ð6Þ
43 M Mass Transport
44
45 and Eq. (2) rewritten as Electrodeposition has the ability to produce a rela-
46 tively uniform distribution of metal upon a cathode
47 ndAhNa e ndAhF of irregular shape. Though the uniformity depends
Q ¼ ¼ ð7Þ
48 M M on the distribution of electric fields inside the electro-
49 lyte toward the surface of the electrode, other impor-
50 where F is Faraday’s constant, defined as F ¼ Na e, tant factors have to be considered. The addition of
51 equal to 96,485.34 C. By rearranging Eq. (7) and know- agents (additives) to the electrolyte, for example, can
52 ing the quantities M, Q, n, d, A and F, the thickness can affect the microscopic mechanism of electrodeposition,
53 be easily calculated from: reducing the roughness of the deposit and producing a
54 visual effect known as brightening.
55 MQ In order to obtain layers with a desired property,
h ¼ : ð8Þ
56 ndAF such as uniform thickness and low roughness, or an
 120037171_E-ECHP_00_00_R2_081905
826 Electrodeposition

1 Helmholtz Layer
A
2 +
+
+
+ +
+
+
+
3 + Bulk Solution

+ +
+
+ +

+ +
+ +
+
+ + +
+ + +
4 + + + +
+
+

+ +
+ +
+
+ +
ELECTRODE

5 +
+
+
+
+
+ +
+
+
+ + +

+ +
+
+ +

+
6 +

+ +
+

+ +

+ +
+

+ +
+ +
+
+ + + +
7 +
+
+
+ +
+ + + + + +
+

+ +
+ + +
8
+ +
+ +

+
+
+ +
+ Solvated+ Ion
+ +
9 +
+
+
+
+ +
+ +
+
+

+ +
+ +
+ +
10 +

+ +
+

+ +
+
+ +
+ +
+ +

+ + +
+ +
11 +
+ +
+
+

water molecule
12
Gouy-Chapman Layer
13 B
14
15 ψSolution Distance from Electrode

16
17
Potential

18
19
20
21
22 ψElectrode Fig. 6 Illustration of the electric
double layer (A) and potential drop AQ1=PR
23 (B) near the surface of the electrode.
24
25
26 electrolyte with high filling capacity, i.e., with the electrolyte and electrode. An electric double layer,
27 ability to produce deposits inside holes or lith ographic illustrated in Fig. 6, will be formed. Fig. 6(A) shows
28 features, one has to consider carefully the transport of the double layer in greater detail. When the exchange
29 different species inside the electrolyte and the reaction process reaches equilibrium the double layer thickness
30 rates of these species on the surface of the electrode. depends on the physical and chemical properties of the
31 Basically, three mechanisms are responsible for electrode=electrolyte interface. In general, the descrip-
32 mass transport inside an electrochemical cell: diffusion, tion of the double layer considers the existence of
33 migration, and convection. Diffusion is mass transport two regions named the Helmholtz layer, a monolayer
34 because of concentration gradients, i.e., variations in of ions on the surface of the electrode, and the
35 the concentration of a species with position. Diffusion Gouy-Chapman layer, a region that penetrates the
36 occurs mainly near the electrode surface because of bulk of the electrolyte with decreasing charge and
37 gradients created by the consumption of species that concentration gradients. In the figure, the ions are
38 undergo redox reactions and are incorporated into shown with a sheath of water molecules. This solvation
39 the deposit. This incorporation process depletes the sheath is because of the electrostatic attraction of polar
40 deposition species near the electrode, generating the water molecules and the ionic species. Fig. 6(B)
41 concentration gradient. illustrates the potential drop near the surface of the
42 The simple introduction of an electrode into an electrode because of the presence of the double layer.
43 electrolyte will lead to an exchange of charge between In the bulk of the electrolyte the potential drop is
44
45
A B
46
47 Electrode Bulk Solution
DIFFUSION d
48 LAYER
49 C
Bulk
50 d'
51
52 C
s' Fig. 7 (A) Idealized profile of the con- AQ1=PR
53 C centration gradient near the surface of
54 Surface the electrode and (B) an illustration of
55 Cs a deposition instability in the presence
d
56 of an asperity.
 120037171_E-ECHP_00_00_R2_081905
Electrodeposition 827

1 A B
2
3 SOLUTION E
4 +
+
+
+
5 + + +
+ +

Solvated Ion
6 +
+ +
+

7
8
9 difusion
growing
10
11 +
+
+ electron
+ difusion over
12 + transfer
+ +
+ +

+
the surface
+
13 + +

14 e– Adatom Cluster Growing

Nucleous
15
16 ELECTRODE
17
18 Fig. 8 (A) Mechanism of formation of electrodeposits on the surface of an electrolyte and (B) an atomic force microscope AQ1=PR
19 (AFM) image reveals the granular nature of a Co deposit. (From Ref.[1].)
20
21 normally very low, because this region is not depleted electrode, as shown in Fig. 7(B). These morphological
22 of ionic species. structures reduce the size of the diffusion layer, thereby
23 Beyond the double layer, there is a depleted region increasing the concentration gradient and hence the
24
named the diffusion layer with a thickness of microns, current. The local increase in current increases the
25
much wider than the double layer, formed during deposition rate, favoring further growth of the asper-
26 deposition by the consumption of a particular species. ities, and consequently an increase in roughness of
27 Fig. 7(A) is a plot of the concentration of an ionic the whole electrodeposited layer. This effect is known
28 species as a function of the distance from the surface as deposition instability.
29 of the electrode, showing the diffusion layer. The Inside the bulk of the electrolyte, mass transport is
30 consumption of ions because of metal deposition gen- mainly because of migration, a mechanism of ionic
31 erates a concentration gradient that, in steady-state motion caused by the presence of an applied electric
32
conditions, is essentially determined by the redox reac- field. In the electrochemical cell the potential drop
33
tion rate. If the consumption of ions arriving at the creates an electric field that is much more intense in
34 surface by diffusion is very high, the concentration of the regions near the surface of the electrodes, but is
35 ions at the surface CS is effectively zero, and the sufficiently intense in the bulk of the electrolyte to
36 deposition process is controlled by diffusion. If the promote the migration of the ions to the border of
37 consumption is low, then the ion concentration at the the diffusion layers.
38 surface is different from zero and the deposition is con- The third important mechanism of mass transport is
39 trolled by kinetics, i.e., by the velocity of the reaction. convection. In this case, the fluid flows in an un-
40
Electrodeposition in the diffusion-limited regime controlled manner because of natural density gradients
41
is very sensitive to asperities on the surface of the (gradients caused by concentration and temperature
42
43
44
A B
45 Metal Deposition Oxide Deposition
46
47 Electrode Solution Al Solution
Electrode
48 2e– Cu2+ 3e– 3H+
49
50
51
Cu
52 Al3+ Al2O3
53
54 Fig. 9 Electrode reaction for (A) the cathodic elec- AQ1=PR
55 trodeposition of Cu and (B) the anodic electro-
Deposit Cu Deposit Al2O3
56 deposition of Al oxide.
 120037171_E-ECHP_00_00_R2_081905
828 Electrodeposition

1
2
3
4
5
6
7
8 Fig. 10 (A) Sequence of filling of
9 a trench profile for the fabrication
10 of Cu interconnects, (B) a cross-
11 section illustration of a six-level
12 wiring structure, and (C) SEM
13 view of IBM’s first-to-market six
14 level copper interconnect technol-
ogy. (From Ref.[2].)
15
16
17
18 fluctuations). Convection can also be produced in a microscopy (AFM) image of an electrodeposited layer
19 controlled manner by different methods, such as from an electrolyte containing cobalt sulfate. The
20 mechanically stirring the electrolyte. image clearly reveals the formation of Co grains on
21 top of electrode surface.
22
23 Growth Mechanisms
24 Electrode Reactions
25 To complete the explanation about the mechanisms of
26 electrodeposition, it is very important to give an idea The current that flows at the working electrode may
27 about the formation of the deposits. A model for the be divided into two kinds: faradaic and non-faradaic.
28 electrodeposition process considers a solvated ion going The faradaic processes are the ones where charges
29 through the diffusion layer as a first step, loosening are transferred across the liquid–solid interface. These
30 of the solvation sheath by transferring electrons processes are called faradaic because they follow
31 with electrode and being adsorbed (adatom) as the Faraday’s law, which says the amount of substance
32 second step, and surface diffusion and incorporation that undergoes oxidation or reduction at each elec-
33 in an energetically favorable site as the third step. trode is directly proportional to the amount of electri-
34 The deposition sites can be punctual or extended sur- city that passes through the cell. Two faradaic
35 face defects, such as vacancies or kinks, known in gen- processes that are directly related to electrodeposition
36 eral as nucleation sites. The nucleation sites allow the are shown in Fig. 9, where Fig. 9(A) represents simply
37 formation of nuclei (islands) that evolve to grains, the deposition of Cu by reduction of Cu2þ and Fig.
38 forming compact deposits that grow on top of the sur- 9(B) the growth of Al anodic oxide by oxidation of
39 face of the electrode. Fig. 8(A) depicts the mechanism metallic aluminum, this being an example of anodic
40 of layer growth and Fig. 8(B) shows an atomic force electrodeposition. Non-faradaic processes are struc-
41
42
43
44 A B
45 0 0
46
Deposition Current (mA)

47
Current (mA)

–2 –2
48 Cu2+ + 2e– = Cu°

49
–4 –4 Fig. 11 (A) Voltammogram of an elec-
50
trolyte containing two salts, Co and Cu AQ1=PR
51 sulfates and (B) pulses corresponding
–6 –6
52 to the alternate deposition of a Cu=Co
53 multilayer. The electrode is semicon-
–8 Co2+ + 2e– = Cu° –8
54 ducting Si. (Courtesy of L. Seligman,
55 –1.2 –0.6 –0.3 –0.0 0 20 40 60 80 100 Universidade Federal de Santa
56 Potential (VxSCE) Time (s) Catarina).
 120037171_E-ECHP_00_00_R2_081905
Electrodeposition 829

1 A B
2 0.6
3 8 Number of E
Layers :
4 0.4 9
6 13
5
AMR %

MR %
19
6 4
0.2
7
2
8
Fig. 12 (A) AMR of single layer of Co[1] and AQ1=PR
9 0.0 0 (B) GMR results depending on the number of
10 –240 –120 0 120 240 –1000 –500 0 500 1000 layers. (From Ref.[5].) The substrate is semi-
11 Applied Field (Oe) Applied Field (Oe) conducting Si.
12
13
14 tural changes of the electrode–solution interface, such ELECTRODEPOSITION IN NANOSCIENCE
15 as absorption and desorption of species that change AND NANOTECHNOLOGY
16 the potential of the electrode and solution composition
17 without charge transfer. Simultaneously, with the rapid growth of electrodepo-
18 sition in microelectronics, a new trend based on the
19 electrodeposition of materials, structures, particles,
20 ELECTRODEPOSITION IN MICROELECTRONICS devices, etc., generally called nano-objects, with
21 dimensions below 100 nm commenced. Nano-objects
22 Recently, there has been a boom in the use of electro- are fundamental for nanoscience investigations and
23 deposition for microelectronics. The microelectronics nanotechnology development. A nano-object is of
24 industry came to the conclusion that the electrodeposi- particular interest if it has physical properties that
25 tion of Cu is the ideal manufacturing process for differ from objects that have macroscopic sizes.
26 wiring for semiconductor logic and memory devices. Quantization of energy, for example, is observed
27 Wiring is the network of wires that interconnects the in systems with greatly reduced size, such as atoms,
28 devices (transistors) on integrated-circuit chips. Cop- molecules, and nanostructures.
29 per is a highly conductive metal and is relatively easy Electrodeposition is an elegant and efficient techni-
30 to electrodeposit. Since 1997, Cu has been successfully que for the production of nano-objects. Using the
31 used for the production of interconnects.[2] Nowadays, pulse deposition mode, it is possible to control the
32 such interconnects are electrodeposited in trenches amount of atoms to be deposited with great precision.
33 with widths of 0.13 mm or less. The ability to fill That is, pulsing with pulse durations of a few millise-
34 trenches and vials with a plated material is called conds to a few seconds, allows the deposition of
35 superfilling and is illustrated in Fig. 10(A). The clusters of atoms or layers with thickness of a few to
36 cross-section of the device illustrated in Fig. 10(B) hundreds of nanometers.
37 depicts the multilevel wiring structure of an integrated A typical example of an electrodeposited nanostruc-
38 circuit and Fig. 10(C) shows a real arrangement of ture is a multilayered structure. By having two salts in
39 electrodeposited interconnects in a device fabricated the electrolyte and applying two potentials in alterna-
40 by IBM. tion, it is possible to deposit multilayer structures,
41
42
43
44 A
45 –4
46
47 –5
48
Current (mA)

49 Fig. 13 (A) Spontaneous current AQ1=PR

–6 oscillations during deposition at
50
constant applied potential using
51 an electrolyte containing Cu
52 –7
sulfate and lactic acid (Courtesy
53 of R.G. Delatorre, Universidade
54 –8 Federal de Santa Catarina) and
55 10 20 30 40 (B) SEM image of Cu=Cu2O mul-
56 Time (s) tilayered wires. (From Ref.[8].)
 120037171_E-ECHP_00_00_R2_081905
830 Electrodeposition

1
2
3 200 200
4
5 0 0
6
Current (mA)

Current (mA)
–200 –200
7
8 –400
–400
9 –600
10 –600
–800
11
–800 0.39 0.40 0.41 0.42
12
Time (s) Fig. 14 Sequence of ultrafast pulses AQ1=PR
13 0.0 0.1 0.2 0.3 0.4 0.5 for the deposition of a nanostruc-
14 Time (s) tured Cu48Ni52 alloy. (From Ref.[10].)
15
16
17 which are artificially fabricated materials that have multilayer. If the individual layers are only a few
18 application in the electronics industry. Fig. 11(A) nanometers thick, which is easily achievable by electro-
19 shows the polarization curve for an electrolyte contain- deposition, the electric resistance will vary with the
20 ing two salts, CuSO4 and CoSO4, where the reduction magnetic field, an effect discovered very recently and
21 peaks of each metal are clearly seen. Pulsing the cath- known as giant-magnetoresistance (GMR).[3,4] The
22 odic potential rapidly between a value at which only Cu first observation of a magnetoresistive effect was by
23 is reduced, and one at which a Co-rich alloy is reduced, Lord Kelvin in 1857 by measuring the electrical resis-
24 generates a deposit that is a Cu=Co multilayer with tance of magnetic alloys. Nowadays, the effect that
25 individual layers of nanometric thickness. The layer he observed is called anisotropic magnetoresistance
26 thickness may be controlled by integrating the current (AMR), and its magnitude does not exceed 6%.
27 in real time and calculating the electrodeposited charge The GMR effect is about one order of magnitude
28 or, more simply, by controlling the deposition time. greater than AMR and depends on many factors, such
29 Fig. 11(B) shows typical current transients with charac- as the thickness and number of individual layers, the
30 teristic peaks for each layer electrodeposited. The Co magnetic material used and the preparation method,
31 deposition current is much higher than the Cu one and is observed also in non-layered granular struc-
32 because to assure the deposition of a Co-rich alloy tures. Magnetoresistive materials have been intensively
33 the concentration of Co sulfate in the electrolyte has used in the high-technology industry as magnetic
34 to be much higher than the Cu sulfate concentration. sensors and reading heads for computer hard disk
35 A multilayer structure with one of the repeating drives. Fig. 12(A) shows the magnetoresistance of a
36 layers being a magnetic material is called magnetic single layer of electrodeposited Co, similar to the one
37
38
39
40 A
41 Pores filled with
42 Nanoporous electrodeposited metal
43 Membrane
44
45
46
47
Metal Coating
48
49 B
50 Cu
51
52 Fig. 15 (A) Alumina membrane with nanopores, AQ1=PR
53 (B) schematic view of a layered nanowire, and
Co-Cu-Ni
54 (C) TEM image revealing the layered structure
55 of a Cu=CuCoNi nanowire grown in alumina
56 nanopores. (From Ref.[11].)
 120037171_E-ECHP_00_00_R2_081905
Electrodeposition 831

1
2
3 E
4
5
6
7
8
9 Fig. 16 (A) High-resolution AFM image of an
atomically flat single crystal of Si showing large
10
terraces and parallel steps and (B) nanowires of
11 Au electrodeposited preferentially at the step
12 edges. (From Ref.[12].)
13
14
15
16 depicted in Fig. 5, which shows an AMR effect of per cycle of 0.25 monolayer of Cu and 0.25 monolayer
17 0.5%. Fig. 12(B) illustrates the case of a magnetic mul- of Ni.
18 tilayer, also electrodeposited, with the effect depending
19 on the number of layers, reaching in this case a
20 maximum of 8.5% for 15 Co=Cu layers. Nanowires
21 An interesting achievement of electrodeposition in
22 the preparation of nanostructures is the self-assembly A special characteristic of electrodeposition is the fact
23 of multilayers.[6] An electrolyte containing copper sul- the deposition occurs only where there is an electrical
24 fate (CuSO4) and lactic acid (C3H6O3) is a standard connection to the external circuit. This is a great
25 example because under certain experimental conditions advantage because it allows the deposition to be area
26 the cell current oscillates spontaneously leading to the selective. By covering the electrode surface with a pat-
27 growth of a nanometric Cu=Cu2O multilayer. In terned insulating layer, electrodeposition will occur
28 Fig. 13(A) the spontaneous oscillations of the deposi- only on the exposed areas. This makes electrodeposi-
29 tion current are illustrated, though a natural damping tion an ideal method for growing materials on
30 of the magnitude is observed. However, in stirred previously determined patterns and also for filling
31 solutions the oscillatory behavior can be maintained high-aspect ratio templates.
32 for several days.[7] Fig. 13(B) shows a SEM image of This advantage can be used for growing nanowires
33 filaments of spontaneously grown Cu=Cu2O multi- (wires with nanometric diameter). Nanoporous mem-
34 layers.[8] branes that can be fabricated by the anodic oxidation
35 The explanation for the spontaneous formation of of aluminum are appropriate templates. This process
36 multilayers lies in variations in the pH.[6,7,9] In the leads to the formation of an alumina layer with parallel
37 growth process, the electrodeposition of Cu2O is nanopores, as shown in Fig. 15(A), which can then
38 favored, as it has an equilibrium potential more positive be filled by electrodeposition. Fig. 15(B) shows a
39 than that of Cu deposition (Table 1). However, the schematic view of a multilayer nanowire and Fig.15(C)
40 reaction 2Cu2þ þ 2e
þ 2OH
$ Cu2 O þ H2 O a transmission electron microscopy image of a Cu=
41 depletes the OH
species near the electrode, locally CuCoNi layered nanowire grown in the nanopores.
42 decreasing the pH and favoring the deposition of Cu. A different way to electrodeposit nanowires is by
43 The above process is repeated, because the OH
con- using the surface of a single crystal as a template.
44 centration is re-established during Cu2þ reduction. Fig. 16(A) shows an AFM image of a silicon surface,
45 Additionally, by having an experimental setup with revealing large terraces with parallel steps. By electro-
46 a high-speed data acquisition system, it is possible to depositing Au at relatively low deposition rates, the
47 control deposition pulses with durations below millise- steps act as deposition sites favoring the formation of
48 conds. This ultrafast pulsing method was called wires of nanometric size along their edges, as shown
49 precision electrodeposition and allowed the deposition in Fig. 16(B).
50 of sub-monolayer quantities of material.[10] Precision
51 electrodeposition was demonstrated for the CuNi
52 system, as shown in Fig. 14, where a sequence of
53 ultrafast current pulses for the electrodeposition of a CONCLUSIONS
54 nanostructured CuNi alloy with a controlled composi-
55 tion of 48% Cu and 52% Ni is displayed. The duration Electrodeposition is a process widely used in industry.
56 of the pulses (tens of milliseconds), allows the deposition In this entry, emphasis was given to fundamental
 120037171_E-ECHP_00_00_R2_081905
832 Electrodeposition

1 aspects and to future potential applications of this cuprous oxide layered nanostructures. J. Mater.
2 technique. Res. 1998, 13 (4), 909–916.
3 7. Switzer, J.A.; Hung, C.-J.; Huang, L.-Y.; Switzer,
4 E.R.; Kammler, D.R.; Golden, T.D.; Bohannan,
5 ACKNOWLEDGMENT E.W. Electrochemical self-assembly of copper=
6 cuprous oxide layered nanostructures. J. Am.
7 The authors wish to thank Prof. Walther Schwarzacher Chem. Soc. 1998, 120, 3530–3531.
8 from Bristol University for reading carefully the manu- 8. Wang, M.; Zhong, S.; Yin, X.-B.; Zhu, J.-M.;
9 script and the Brazalian agencies CAPES, CNPQ, and Peng, R.-W.; Wang, Y.; Zhang, K.-Q.; Ming, N.-B.
10 FAPESC. Nanostructured copper filaments in electroche-
11 mical deposition. Phys. Rev. Lett. 2001, 86, 3827–
12 3830.
13 REFERENCES 9. Zhang, M.-Z.; Wang, M.; Zhang, Z.; Zhu, J.-M.;
14 Peng, R.-W.; Ming, N.-B. Periodic structures of
15 1. Munford, M.L.; Sartorelli, M.L.; Seligman, L.; randomly distributed Cu=Cu2O nanograins and
16 Pasa, A.A. Morphology and magnetic properties periodic variations of cell voltage in copper elec-
17 of Co thin films electrodeposited on Si. J. Electro- trodeposition. Electrochim. Acta 2004, 49 (14),
18 chem. Soc. 2002, 149 (5), C274–C279. 2379–2383.
19 2. Andricacos, P.C. Copper on-chip interconnec- 10. Kazeminezhad, I.; Blythe, H.J.; Schwarzacher, W.
20 tions: a breakthrough in electrodeposition to Alloys by precision electrodeposition. Appl. Phys.
21 make better chips. Interface (Electrochem. Soc.) Lett. 2001, 78 (7), 1014–1016.
22 1999, 8, 33–48. 11. Schwarzacher, W. Metal nanostructures: a new
23 3. Baibich, M.N.; Broto, J.M.; Fert, A.; Nguyen Van class of electronic devices. Interface (Electrochem.
24 Dau, F.; Petroff, F.; Eitenne, P.; Creuzet, G.; Soc.) 1999, 8, 20–24.
25 Friederich, A.; Chazelas, J. Giant magnetoresis- 12. Munford, M.L.; Maroun, F.; Cortes, R.;
26 tance of (001)Fe=(001)Cr magnetic superlattices. Allongue, P.; Pasa, A.A. Electrochemical growth
27 Phys. Rev. Lett. 1988, 61, 2472–2475. of gold on well-defined vicinal H–Si(1 1 1) sur-
28 4. Binasch, G.; Grünberg, P.; Saurenbach, F.; Zinn, faces studied by AFM and XRD. Surf. Sci.
29 W. Enhanced magnetoresistance in layered mag- 2003, 537 (1–3), 95–112.
30 netic structures with antiferromagnetic interlayer
31 exchange. Phys. Rev. B. 1989, 39, 4828–4830.
32 5. Seligman, L.; Sartorelli, M.L.; Pasa, A.A.;
33 Schwarzacher, W.; Kasyutich, O.I. Co=Cu spin FURTHER READINGS
34 valves electrodeposited on Si. J. Magn. Magn.
35 Mater. 2001, 226, 752–753. Brad, A.J.; Faulkner, L.R. Electrochemical Methods:
36 6. Switzer, J.A.; Hung, C.J.; Huang, L.Y.; Miller, Fundamentals and Applications; Wiley: New York,
37 F.S.; Zhou, Y.C.; Raub, E.R.; Shumsky, M.G.; 1980.
38 Bohannan, E.W. Potential oscillations during Schlesinger, M., Paunovic, M., Eds.; Modern
39 the electrochemical self-assembly of copper Electroplating, 4th Ed.; Wiley: New York, 2000.
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56