Growth modes of electrodeposited cobalt

Article abstract of DOI:10.1016/j.electacta.2004.04.001

Structure, morphology and electrokinetics in cobalt electrocrystallisation from sulphamate electrolytes, either without or with addition of boric acid, are studied and an interpretation of the structure characteristics and growth modes of electrodeposited cobalt is proposed. The correlation between crystallographic structure and electrokinetic behaviour, analysed by short galvanostatic pulses over-imposed during steady-state growth at fixed current density, is investigated, pointing out the main factors which affect structure and texture of electrodeposits and emphasising the electrolyte pH influence. Three well-defined growth modes are recognised and univocally related to the transient electrokinetic parameters. Growth structures characterised by low crystallinity and/or texture modifications occur in relation with the disturbance of the hydrogen discharge reaction or as a result of ageing of not-buffered electrolyte. The basic growth modes are classified as follows: outgrowth or perpendicular growth, in condition of surface stabilisation of hydrolysed species resulting in low nucleation activity and growth of relatively large-grained columnar deposits; lateral growth, when hydrolysed species are stable in the bulk solution and cellular crystallisation occurs as a consequence of full surface coverage by discharge intermediates and precipitation of insoluble hydrolysis products at grain boundaries; cluster growth, in the presence of boric acid, in conditions of nucleation control, possibly related to the formation of adsorbed complexes stabilised at the surface by boric acid.

Full text of DOI:10.1016/j.electacta.2004.04.001

Electrochimica Acta 49 (2004) 4079–4089
Growth modes of electrodeposited cobalt
A. Vicenzo∗,1, P.L. Cavallotti1
Dipartimento di Chimica, Materiali, Ingegneria Chimica “Giulio Natta”, Politecnico di Milano,
Via Mancinelli, 7-20131 Milano, Italy
Received 8 November 2003; received in revised form 1 April 2004; accepted 2 April 2004
Available online 8 May 2004
Abstract
Structure, morphology and electrokinetics in cobalt electrocrystallisation from sulphamate electrolytes, either without or with addition
of boric acid, are studied and an interpretation of the structure characteristics and growth modes of electrodeposited cobalt is proposed.
The correlation between crystallographic structure and electrokinetic behaviour, analysed by short galvanostatic pulses over-imposed during
steady-stategrowthatfixedcurrentdensity, isinvestigated, pointingoutthemainfactorswhichaffectstructureandtextureofelectrodepositsand
emphasising the electrolyte pH influence. Three well-defined growth modes are recognised and univocally related to the transient electrokinetic
parameters. Growth structures characterised by low crystallinity and/or texture modifications occur in relation with the disturbance of the
hydrogen discharge reaction or as a result of ageing of not-buffered electrolyte.
The basic growth modes are classified as follows: outgrowth or perpendicular growth, in condition of surface stabilisation of hydrolysed
species resulting in low nucleation activity and growth of relatively large-grained columnar deposits; lateral growth, when hydrolysed species
are stable in the bulk solution and cellular crystallisation occurs as a consequence of full surface coverage by discharge intermediates and
precipitation of insoluble hydrolysis products at grain boundaries; cluster growth, in the presence of boric acid, in conditions of nucleation
control, possibly related to the formation of adsorbed complexes stabilised at the surface by boric acid.
© 2004 Elsevier Ltd. All rights reserved.
Keywords: Cobalt electrodeposition; Electrodeposits structure; Electrodeposits texture; Electrokinetics; Iron group metals
1. Introduction
mechanism participating in Co electrocrystallisation is ex-
pected to contribute to the systematic understanding of the
structure–properties correlation of ECD Co and Co alloys,
as stressed in a recent review [7].
The easy occurrence of preferred orientation (PO) in elec-
trodeposited (ECD) iron group metals has stimulated a num-
ber of investigations, without, however, achieving the stage
of a comprehensive view. A general and exhaustive interpre-
tation, centred on Fischer’s concept of inhibition [1], was
proposed for nickel electrocrystallisation from Watts’s elec-
trolyte [2,3] and later extended to cover the different texture
types from simple sulphate and chloride solutions [4,5].
Cobalt electrocrystallisation has been far less studied
compared to Ni, although ECD Co and Co alloys are el-
igible materials for applications ranging from magnetic
media and devices [6] to wear and corrosion resistant coat-
ings. Advances in the comprehension of the factors and
Electrolytic Co crystallises with both hexagonal closed
packed (hcp, ␣-Co), the stable allotropic modification at tem-
perature below 417 ◦C, and face centred cubic (fcc, ␤-Co)
lattice structure, as first reported by Hull [8]. The solution
pH was shown to be the most important parameter in deter-
mining the structure of ECD Co by Kersten [9]. Pangarov
and Rashkov [10] determined the relations between the main
operative conditions and electrodeposits structure: ␤-Co for-
mation is favoured by low temperature, high current density
(cd) and low pH. Polukarov [11] came to similar conclu-
sions. These early results were confirmed by several authors
[12–14].
The formation of ␤-Co by electrodeposition was studied
in details and still remains a research topic of primary rele-
vance, although in recent years the attention has been partic-
ularly focused on phase selection as related to substrate in-
∗
Corresponding author. Tel.: +39-02-2399-3124;
fax: +39-02-2399-3180.
E-mail address: antonello.vicenzo@polimi.it (A. Vicenzo).
1
ISE member.
0013-4686/$ – see front matter © 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2004.04.001

4080
A. Vicenzo, P.L. Cavallotti / Electrochimica Acta 49 (2004) 4079–4089
fluence [15,16] and deposition conditions [15,17,18] rather
than on the mechanism of phase formation, for which the
reference papers are still those of Gaigher and Van der Berg
[19] and of Nakahara and Mahajan [20].
can be rationalised on the basis of pH and boric acid influ-
ence [36]. In the present paper the structure–kinetics correla-
tion is analysed in details and explained considering the role
of the different electrolyte species involved in the growth
process at the cathode.
The first extensive reports on the crystal structure of ECD
␣-Co appeared in 1966, by Cadorna and Cavallotti [12]
and by Pangarov and Vitkova [21]. According to the lat-
ter, the experimental results were in substantial agreement
with the theory of two-dimensional nucleation [22]. The fol-
lowing investigations were almost exclusively focused on
deposition from Watts’s type electrolyte or from buffered
sulphate solution. Deposits having two-phase composition
show a hcp [11.0] + fcc [1 1 0] texture, from buffered sul-
phate [13,21,23,24] or sulphamate [12] solution, in agree-
ment with the early observation of Finch et al. [25]. With
increase of pH the structure becomes completely of the
␣-phase and deposits texture depends mainly on solution pH
[12,23–26] and, to a minor extent, on temperature [23] and
cd [27]. The prevailing orientations and their stability with
respect to the operative conditions are characterised in de-
tails for both sulphate [12,13,23,24,28] and chloride based
electrolytes [29].
Early attempts to interpreter the growth modes of ECD
Co followed Reddy’s views on Ni electrocrystallisation [30].
The different growth textures were explained as consequence
of competitive adsorption between hydrogen and intermedi-
ate species [23] or classified according to growth conditions
related to different degree of inhibition [13]. Subsequently,
the main importance of the electrolyte pH was recognised.
It was suggested [20] or explicitly stated [19] that the sur-
face pH and the related chemical phenomena at the cath-
ode surface were the major factors responsible for the de-
velopment of different textures [26,31]. The structural bear-
ing of hydrolytic phenomena received further support from
the work of Croll [26], who reported that, depositing from
CoSO4 electrolyte, in the absence of H3BO3 a sharp change
from [10.0] + [11.0] to [00.1] PO takes place at pH about
4 within a short range, while when H3BO3 is present, the
[10.0] + [11.0] PO weakens gradually as the pH raises and
the transition pH is higher, about 6.4. The possibility of ob-
taining the three main growth textures of ␣-Co by autocat-
alytic deposition [32] and the assessment of the conditions
giving Co cellular growth [33,34] with strong [00.1] texture
definitely pointed out the decisive contribution of reactive
hydrolysis in determining the deposits structure.
2. Experimental
The plating solutions were prepared from analytical grade
chemicals and distilled water. Electrolytes with 1 M Co2+
concentration were prepared from Co carbonate and sul-
phamic acid (in the following, the symbol ␴ stands for
the sulphamate anion, H2NO3S−). H3BO3, in concentra-
tion 0.4 or 0.5 M, CoCl2 and sulphamid were added in
some instances as outlined in Table 1, where composition
and operation conditions for all plating baths are listed.
The electrolytes were pre-treated to eliminate organic im-
purities with active charcoal (1 h, 80 ◦C); in addition, some
solutions were pre-treated with electrolysis at low current
(cd < 1 mA cm−2) and with large cathodic area. Since
no detectable differences were noticed between same baths
before and after this treatment, it was thereafter omitted.
Plating baths pH was adjusted by addition of either cobalt
hydroxy-carbonate or sulphamic acid. The substrates were
either brass sheets coated with about 3 ␮m ACD NiP or brass
sheets mechanically polished with 1200 grit emery-paper.
Either a dilute HCl (NiP coated brass) or H␴ (brass) etch,
followed by washing with distilled water and drying with
nitrogen gas, was used immediately before plating. High pu-
rity Co was used as anode.
Phase structure and texture of deposits were determined
by X-ray diffractometry (XRD) with Cu K␣ radiation and a
powder goniometer. All electrodeposits were thick enough
to show the crystallographic orientation imposed by the
bath (not less than 20 ␮m). In order to characterise texture
modification of ␣-Co deposits with electrolyte composition
(namely, pH in the range 2.7–6 and boric acid addition), the
volume fraction Mhkl of different texture components was
estimated by the equation [37,38]:
Fhkl
ꢀ
Mhkl
=
hkl Fhkl
From the standpoint of the phenomenological comprehen-
sion, the literature about Co electrodeposition provides the
elements for understanding the relationship between opera-
tive conditions and crystallographic structure of electrolytic
Co. On the ground of this well-established knowledge, the
main objective of the present work is to investigate the corre-
lation between structure and kinetics in Co electrocrystalli-
sation from sulphamate solutions. We have already shown
that the electrokinetic behaviour is influenced by hydrolysis
conditions at the surface [35] and that the structure–kinetics
relationship in Co electrodeposition from sulphamate bath
Table 1
Composition and operative conditions of the electrolytes (NH2SO3: ␴)
Solution
pH
T (◦C)
cd (mA cm−2
)
Co␴ 1 M
1.8–6
6
30
30
4–40
5–20
2
Co␴ 1 M–H4N2O2S
2
(sulphamid) 5 mM
Co␴ 1 M–H3BO3 0.4 M
2.5–4.8
4.3
30
30
10
10–20
2
Co␴ 1 M–H3BO3
2
0.5 M–CoCl2 0.1 M
Co␴ 1 M–H3BO3
5.2
30
10
2
0.5 M–CoCl2 0.1 M

A. Vicenzo, P.L. Cavallotti / Electrochimica Acta 49 (2004) 4079–4089
4081
where the texture index Fhkl for( h k l) plane is given by the
expression:
transients; a linear relationship between ηω and (iP − iD)
was found in all instances.
I
hkl/IhRkl
Fhkl
=
ꢀ
(1/n) hkl Ihkl/IhRkl
3. Results
Ihkl is the measured intensity of reflection (h k l); IhRkl is the
reflection intensity of a random powder sample; and n is
the number of measured reflections. This procedure is valid
on the assumption that only crystals with orientation along
the measured directions occur, that is, (10.0), (00.2), (10.1),
(11.0); and that a low or nil volume fraction of ␤-Co is
present so as not to contribute appreciably to the reflection
intensities of the (00.2) and (11.0) lines.
Surface morphology of thick deposits was investigated
by scanning electron microscopy (SEM). Atomic force mi-
croscopy (AFM) in conjunction with glancing angle XRD
analysis was used to characterise the initial stages of growth
of Co thin films electrodeposited from pH 4 buffered and
not-buffered solution at 10 mA cm−2 and 30 ◦C. Thin films
of up 1 ␮m thickness were deposited on Siꢀ1 0 0ꢁ/TiN/Cu
substrate; the Cu seed layer was 100 nm thick and [1 1 1]
textured.
Potentiodynamic, steady-state and transient galvanostatic
experiments were carried out in a standard three-electrode
configuration, with either an Ag/AgCl reference elec-
trode (RE) in connection to the working electrode with a
lateral-channel Piontelli probe [39] (potentiodynamic and
steady-state galvanostatic measurements) or a Pt RE (tran-
sient measurements). A Model 273A EG&G PAR and an
AMEL System 5000 potentiostat/galvanostat were used. Po-
tentiodynamic runs (scan rate 0.1 mV s−1) and constant cur-
rent potentiometry were performed on electrodes previously
coated (about 3 ␮m) in the same electrodeposition bath.
No ohmic correction was made, when using the Piontelli
probe.
The kinetic behaviour during deposition was investigated
with the secondary current pulse (SCP) technique [40,41].
The principle of the SCP method is that of a direct current
relaxation technique in which the electrode, which is kept in
a steady-state at fixed deposition cd, is perturbed by a cur-
rent pulse and the relaxation transient is recorded and anal-
ysed by non-linear fitting procedure in order to estimate the
parameters of the equivalent circuit of the electrode system,
as described later in this article.
3.1. Structure and morphology of ECD Co from
not-buffered solutions
Structure and texture of Co deposited from Co␴ 1 M
2
solution, at 30 ◦C and cd in the range up to 40 mA cm−2, is
mainly influenced by the solution pH.
In the acidic region at pH < 2.5, deposits crystal struc-
ture is characterised by the presence of both ␣ and ␤-Co,
with high density of stacking faults and lattice strain, as
shown by (10.0) peak asymmetry [42] and by peaks broad-
ening [43]. In Fig. 1A, the surface morphology of a 20 ␮m
thick coating (pH 1.8, cd 40 mA cm−2) shows coexistence
of different features related to the two-phase structure: elon-
gated ridge-shaped crystals embedded in a fine-grained ma-
trix. The transition from ␤ to ␣-Co occurs in the pH range
2.5–3.0. Coatings obtained in this pH range show a faulted
hcp crystallographic structure, characterised by (10.1) line
broadening [44], low crystallinity and a faint two-fold [11.0]
and [10.0] PO; no formation of ␤-Co in appreciable amount
occurs in this pH range, according to XRD analysis, that is
the (2 0 0) cubic diffraction line is either very weak, for the
low pH values in the above range, or not observed; however,
even in the later case, the presence of a low volume fraction
of ␤-Co, as a finely disperse phase, cannot be ruled out.
Increasing the electrolyte pH, a transition occurs from the
two-fold [11.0] and [10.0] PO to a strong [11.0] PO at pH
3.2( 0.2). Deposits with [11.0] PO are obtained at pH in
the range 3.2–5.7 and cd 5–40 mA cm−2. The surface mor-
phology of ␣-Co deposits with strong [11.0] PO is shown in
Fig. 1B and C and Fig. 2A and B. The characteristic features
change with pH, from dihedral grains, resulting from piling
up of outgrowing basal planes, in the more acidic range; to
pyramidal grains of four or five-fold symmetry at higher pH,
of size 0.5–2 ␮m, increasing as pH decreases. This charac-
teristic crystal habit is shown in Fig. 2A and B by surface
SEM micrographs on 20 ␮m thick coatings electrodeposited
from a pH 4 Co␴2 1 M electrolyte at 10 mA cm−2 and 30 ◦C,
either without or with the addition of Cl− 0.2 M, respec-
tively. Growth of pentagonal pyramids is a consequence of
multi-twinned crystallite formation, as already observed for
Ag [45,46] and Ni [47,48] fcc metals. Similar morphologi-
cal features were reported by Weil and co-workers [13] and
Winand and Scoyer [29] for [11.0] textured deposits.
SCP measurements were performed over-imposing short
(4–6 ms) square current pulses ranging from (iD + 10) to
(iD + 50) mA cm−2 during deposition at iD cd. Before mea-
surements, cobalt was deposited from the bath under study
at iD for 30 min. Pulses were applied every 3–5 min time
interval depending on the bath; interface and growth stabil-
ity was checked by monitoring the cathode potential dur-
ing deposition. SCP samples were always subjected to XRD
analysis and SEM observation showing no differences with
respect to samples prepared in standard conditions. Ohmic
drop in solution, ηω, could be readily determined from the
The observed correlation between electrolyte pH and
[11.0] PO is influenced by electrolyte ageing in the labora-
tory environment at room temperature. Ageing results in the
suppression of the [11.0] PO and the formation of a faint
two-fold [10.0] and [11.0] texture, which corresponds to the
growth structure observed depositing from electrolytes at
pH below about 3.2; eventually, coatings structure changes

Fig. 1. Surface SEM micrographs of Co deposits from Co␴ 1 M, at 30 ◦C for 2 h. (A) pH 1.8, 40 mA cm−2 (20k×); (B) pH 3.2 (20k×); and (C) pH 4.5 (50k×), both at cd 10 mA cm−2
.
2

A. Vicenzo, P.L. Cavallotti / Electrochimica Acta 49 (2004) 4079–4089
4083
Fig. 2. Surface SEM micrographs of Co deposits from Co␴ 1 M, at cd 10 mA cm−2, 30 ◦C, for 2 h. (A) pH 4.5 (50k×); (B) +CoCl2 0.1 M, pH 4
2
(30k×); and (C) +H4N2O2S 5 mM, pH 6 (50k×).
into a weak [00.1] PO and may further evolve towards
a quasi-amorphous state. According to our observations,
these changes occur over a variable time span up to several
months. An example of such structure evolution is presented
in Fig. 3, where XRD patterns of Co deposits from a sim-
ple Co␴2 1 M solution at 10 mA cm−2 and 30 ◦C are shown.
A freshly prepared electrolyte of pH 5 gives deposits with
[11.0] PO (pattern A in Fig. 3); at pH 3.3 (pattern B) de-
posits with the two-fold [11.0] and [10.0] PO; the same so-
lution, after 6 weeks ageing at room temperature and pH 3.3,
gives coatings showing a highly defective [00.1] P.O. (pat-
tern C). This time-variant behaviour is apparently related to
the evolution of hydrolysis reactions in the plating solution,
resulting in the formation of hydroxyl complexes, polynu-
clear or even polymeric species [49]. The attribution of so-
lution ageing to hydrolytic phenomena is supported by the
observation that 0.1 M CoCl2 addition to pH 4 electrolyte
greatly stabilises the [11.0] textured growth mode, which is
practically not affected by ageing time: [11.0] textured coat-
ings were obtained from a bath of such composition over a
period of 24 months. The vanishing of ageing effects in the
presence of chloride can be readily understood as a result
of chloro–cobalt and/or chloro–hydroxy–cobalt complexes,
which are expected to appear as stable species at such Cl−
concentration, similarly to the case of nickel [50], thus influ-
encing the hydrolytic speciation in the Co sulphamate elec-
trolyte.
This time dependent behaviour related to slow hydrolysis
reactions affects also deposits nucleation: after six months
ageing time, a pH 4 solution gives strong [00.1] PO on
100 nm thick [1 1 1] textured Cu film, sputtered on Si wafer;
on the other hand, a heavily defective [00.1] structure is
obtained on brass coated with ACD NiP.
The [00.1] PO is usually obtained from pure sulphate or
sulphamate solutions at high temperature and pH, with some
misorientation shown by the appearance of the (10.1) reflec-
tion. A [00.1] texture at low temperature and pH 5 was pre-
viously reported by Pangarov as an exception to the theory
of two-dimensional nucleation [21]. According to previous
findings [33], the [00.1] texture can be obtained from pure
Co␴ 1 M solution at pH 6 with high perfection, in condi-
2
tions of cellular growth, that is directional crystallisation of
needle shaped crystallites with basal plane parallel to the
surface, isolated one from the other by precipitated hydrox-
ides or basic salts. The addition of sulphamid 5 mM strongly
stabilises this growth texture, giving bright and smooth de-
posits with very high [00.1] PO, even at room temperature.
Surface morphology of these deposits is shown in Fig. 2:
grains are very small and roundish, with 50–100 nm diame-
ter; similar values for the crystallite size, in the range from
80 to 100 nm, are derived from (00.2n) lines broadening ac-
cording to the Williamson–Hall method [51], as expected
for the condition of cellular growth.
XRD patterns representative of the structure of cobalt
coatings electrodeposited from not-buffered Co␴ elec-
2
trolytes at changing pH are shown in Fig. 4. The texture
change occurring in ␣-Co deposits from not-buffered solu-
tions as the electrolyte pH is varied in the range 2.7–5.7 is
summarised in Fig. 5A by the change of the [h k l] orienta-
tion volume fraction.
3.2. Structure and morphology of ECD Co from buffered
solutions
Co deposits from simple solutions buffered with boric
acid, Co␴ 1 M, H3BO3 0.4 or 0.5 M, at pH 3.8–4.8, 30 ◦C
2
and 10 mA cm−2, have a characteristic two-component
[11.0] + [10.0] texture, in the following referred to as
“mixed texture” [27]. XRD patterns show strong reflections
for (10.0) and (11.0) planes and a weak (20.1) peak; (00.2)
and (10.n) lines are absent, contrary to the case of Co coat-
Fig. 3. XRD pattern of Co deposits from Co␴ 1 M (10 mA cm−2, 2 h,
2
30 ◦C). (A) freshly prepared solution, pH 5; (B) idem, pH 3.3; and (C)
after 6 weeks ageing, at pH 3.3.

4084
A. Vicenzo, P.L. Cavallotti / Electrochimica Acta 49 (2004) 4079–4089
Fig. 7. XRD patterns of Co coatings electrodeposited on brass/NiP sub-
strate at 10 mA cm−2 and 30 ◦C from Co␴ 1 M, H3BO3 0.4 M. (A) pH
2
3.6; (B) pH 3.2; (C) pH 2.7; and (D) pH 1.8.
Fig. 4. XRD patterns of Co coatings electrodeposited on brass/NiP sub-
strate at 10 mA cm−2 (pH from 2.5 to 6) and 40 mA cm−2 (pH 1.8) and
30 ◦C from Co␴ 1 M electrolytes at varying pH. The deposition time
and B). The characteristic morphology of the mixed tex-
ture was described by Froment and co-workers [27,52]: the
two symmetrical halves of the ridged grains were shown
by electron diffraction to be differently oriented, one with
[10.0] axis, the other with [11.0] axis normal to the base.
2
was 2 h at 10 mA cm−2 and 45 min at 40 mA cm−2
.
Fig. 7 presents XRD patterns of Co deposits from Co␴
2
1 M, H3BO3 0.4 M and pH 1.8–3.6. A few significant mod-
ifications of the mixed structure should be pointed out: the
(10.1) peak appears at pH < 3.5; the (00.2) reflection in-
tensity increases with decreasing pH below 3.2, with evi-
dence of ␤-Co formation at pH < 2.8, as shown by the
appearance of the (2 0 0) cubic line in XRD patterns. Cor-
respondingly, as shown in Fig. 5B, the volume fraction of
the [10.0] orientation decreases, whilst that of the [11.0] ori-
entation increases, also as a possible spurious contribution
from the [1 1 0] ␤-Co component. These structural changes
are suggestive of the progressive formation of ␤-Co as pH
decreases, through an intermediate region where accumula-
tion of growth defects initiate the ␣ to ␤-Co structure tran-
sition. In particular, the appearance of the (00.2) + (1 1 1)
reflection at pH below 3.2 points to the destabilising effect
of the hydrogen evolution reaction on the mixed structure,
which can accommodate most compact planes only normal
to the surface.
Fig. 5. Change of the [h k l] orientation volume fraction for ECD ␣-Co
vs. the electrolyte pH. Co deposition from Co␴ 1 M (A) and from Co␴
2
2
1 M, H3BO3 0.4 M solution and (B) at 10 mA cm−2 and 30 ◦C.
ings from not-buffered electrolytes with the above-called
two-fold [10.0] + [11.0] PO. The surface morphology is
characterised by randomly arranged 2–5 ␮m long ridges,
possibly curved, consisting of fine sub-grains (see Fig. 6A
Fig. 6. Surface SEM micrographs of Co deposits from Co␴ 1 M, H3BO3 0.4 M, after 2 h deposition at 10 mA cm−2, and 30 ◦C. (A) pH 4 (20k×); (B)
2
pH 4.8 (20k×); and (C) +CoCl2 0.1 M, pH 5.2 (20k×).

A. Vicenzo, P.L. Cavallotti / Electrochimica Acta 49 (2004) 4079–4089
4085
Fig. 10. Three-dimensional AFM images of the surface topography of Co
thin layer on Si/TiN/Cu substrate, after deposition for 60 (left) and 300 s
(right) from a pH 4, Co␴ 1 M solution at 30 ◦C and 10 mA cm−2
.
2
AFM surface topography gives further support to XRD
investigations. Fig. 10 shows AFM images of Co thin films
obtained from the aged not-buffered electrolyte, after 60 and
300 s deposition time. These films, obtained with 94% cur-
rent efficiency, are discontinuous even at 200 nm thickness
and characterised by irregularly sized and shaped grains, re-
sulting from the coalescence of smaller crystallites. After
300 s deposition, grain size increases and grain boundaries
appear well defined around single grains in relief, while the
outer projection of single grains remains unchanged, com-
pared to the 200 nm thick layer.
Fig. 8. Glancing angle XRD patterns of Co deposits from Co␴ 1 M
2
(lower curve) and Co␴ 1 M, H3BO3 0.4 M (upper curve) on Si/TiN/Cu
2
substrate, at pH 4, 10 mA cm−2, and 30 ◦C, after 60 and 120 s deposition
time, respectively (S: substrate).
Deposits obtained from solutions containing also CoCl2
0.1 M, at pH 4.3 or 5.2 show the same mixed structure
with the [11.0] orientation component prevailing as cd in-
creases. Surface morphology, as shown in Fig. 6C, is slightly
changed, consisting of smaller ridges, the size of 2 ␮m or
less, either straight or bent, often gathered in thick bundles.
In Fig. 11, AFM images of thin Co films from Co␴ 1 M,
2
H3BO3 0.4 M, pH 4 solution, are shown. After 10 s deposi-
tion time, at 10 mA cm−2 and 30 nm nominal thickness, the
surface shows small crystallites the size of 100 nm or less,
forming a continuous layer. The morphology characteristic
of the mixed texture, with prismatic grains in the shape of
ridges, is already observed in this early stage of growth.
After 20 s deposition, at the same cd and with 60 nm nom-
inal thickness, AFM topography shows a continuous sur-
face of similar morphological features; grain size is only
slightly increased. After 300 s deposition, at 1 ␮m nominal
thickness, the characteristic surface morphology of [10.0]
+ [11.0] mixed texture deposits is definitively shaped.
3.3. Early stages of growth
Fig. 8 shows glancing angle XRD patterns of Co thin
deposits from Co␴2 1 M solutions, in the absence or presence
of H3BO3 0.4 M, on Si/TiN/Cu [1 1 1], at pH 4, 10 mA cm−2
,
and 30 ◦C, after 60 and 120 s deposition, respectively. From
the 6-month-aged not-buffered solutions, a [00.1] texture is
already emerging, also as a result of epitaxial growth on the
Cu [1 1 1] substrate; from the buffered solution, the mixed
texture is obtained, despite the adverse substrate influence.
After 300 s deposition, as shown in Fig. 9 by XRD patterns of
Co deposits on the same substrate and in the same operative
conditions, a strong [00.1] PO is observed from the aged
not-buffered solution, while from the buffered solution the
mixed texture [10.0] + [11.0] prevails.
3.4. Electrokinetic behaviour
The SCP transient technique was used to characterise in-
terface phenomena during Co deposition from buffered and
not-buffered solution at changing pH.
The overvoltage transient during SCP measurements is
described with a two terms equation: the first resulting from
the equivalent circuit of the potentiometric cell (consisting
of the parallel of a non-linear resistor with Tafel character-
istic and a variable adsorption pseudo-capacitance); the sec-
ond from a linearised Sand-type contribution. The following
equation of the overpotential transient results [40,41]:
ꢅ
ꢁ
ꢂ
ꢃꢄ
iP
iP − iD
iDt
RT
zF
t
η = BT ln
−
exp −
+
iD
iD
BTCads
τ
(1)
with parameters: BT (mV dec−1) the transient Tafel slope,
Cads (␮F cm−2) the adsorption pseudo-capacitance, τ (ms)
the relaxation time; iD the deposition cd and iP the pulse cd.
BT is related to the asymptotic value reached after charging;
Fig. 9. XRD patterns of Co deposits from Co␴ 1 M (upper curve) and
2
Co␴ 1 M, H3BO3 0.4 M (lower curve) on Si/TiN/Cu substrate, at pH 4,
2
10 mA cm−2, and 30 ◦C, after 300 s deposition time (S: substrate).

4086
A. Vicenzo, P.L. Cavallotti / Electrochimica Acta 49 (2004) 4079–4089
Fig. 11. Three-dimensional AFM images of the surface topography of Co thin layer on Si/TiN/Cu from a pH 4, Co␴ 1 M, H3BO3 0.4 M, solution
2
(30 ◦C and 10 mA cm−2). From left to right: 10, 20 and 300 s deposition time.
if a steady overvoltage value is not reached at the end of the
pulse, when phenomena preceding or parallel to the charge
transfer step influence the cathodic process, BT can be in-
ferred from the time behaviour of the transient, introduc-
ing the second term for the overpotential. The capacitance
behaviour at the electrodic surface is related to the nature
and amount of adsorbed electroactive species and the ex-
pression Cads exp(η/BT)(dη/dt) is assumed for the capaci-
tive cd, accounting for the influence of the surface overpo-
tential η on the pseudo-capacitance change at the electrode.
With regard to the t1/2 term, neither diffusive nor convec-
tive effects are likely to affect the transient behaviour at the
electrode in a time length as short as 6 ms, with high con-
centration of reactants and slow electrode reactions; then,
the SCP method, in our experimental conditions, can effec-
tively separate the surface phenomena from mass-transport
effects. Consequently, except for a few cases, namely when
the hydrogen discharge reaction plays an important role as a
side-reaction, the t1/2 term is representative of adsorption ef-
fects involving the modification of the surface concentration
of either reactant or inhibiting species [53]. In fact, a linear
relationship between τ1/2 and the reciprocal of (iP −iD) was
observed, while the product of τ1/2 times iP or (iP −iD) was
not constant, showing that the observed phenomena do not
conform to the case of electrode process controlled by dif-
fusion [54]. Fitting of the experimental data was carried out
by a non-linear least-square optimisation procedure; a good
agreement was always found between experimental and the-
oretical transient, with Chi-square values as a rule in the
range 0.2–0.01.
Fig. 12 shows the influence of pH upon SCP parameters
for Co␴ 1 M solution and Co␴ 1 M, H3BO3 0.4 M so-
2
2
lutions. The reported values are the average of 5–10 mea-
surements at changing pulse cd in the range from 20 to
60 mA cm−2; the error bar refers to the difference between
the average and the highest and lowest value obtained.
For simple sulphamate solutions, the SCP parameters BT
and Cads are similarly influenced by pH change. In the acidic
region BT is in the range 2RT/F, then decreases abruptly in
the pH range 3–3.4 and levels to 70 mV dec−1 at higher pH;
Cads starts from about 100 ␮F cm−2 at low pH, depending
on the rate of hydrogen evolution, becoming minimum at
pH around 3.2 and raising slowly at higher pH. The tran-
sient shape is affected in a peculiar fashion: at pH < 3, the
transient overvoltage does not reach a steady value at low
pulse cd (iP from 2 to 4 times iD) and the t1/2 term is in-
troduced to account for the observed polarisation effect; τ
values range from 50 to 100 ms. This behaviour can be re-
Fig. 12. SCP parameters BT and Cads vs. pH for Co␴ 1 M solution (on the left side of the graph) and for Co␴ 1 M, H3BO3 0.4 M solution (on the right).
2
2
Reported values are average of 5–10 measurements at different pulse cd iP 20, 30, 40, 60 mA cm−2, and deposition cd iD 10 mA cm−2 (40 mA cm−2 at
pH 1.8), at 30 ◦C.

A. Vicenzo, P.L. Cavallotti / Electrochimica Acta 49 (2004) 4079–4089
4087
lated to adsorbed hydrogen disturbance, which fades away
4. Discussion and conclusion
as the pulse cd and pH are raised. At pH > 3.2, the transient
shape changes completely: the charging time becomes very
short, resulting in low Cads values down to 30 ␮F cm−2, and
the transient overvoltage goes through a slight maximum.
With increasing pH, the initial maximum disappears and a
well-defined steady overvoltage is reached in less than 2 ms.
Chloride addition causes a small increase of Cads and BT.
Aged (6 months from preparation) not-buffered sulpha-
mate solutions at pH 3.2–3.6 shows a transient kinetic be-
haviour thoroughly different from that of not-aged baths of
same pH. The transient shape reproduces that observed in
not-aged bath at pH < 3. SCP parameters remain almost un-
changed with pH: BT in the range of 100 mV dec−1 and Cads
about 45 ␮F cm−2. These values can be compared with those
of not-aged electrolyte at pH < 3: BT 110( 5) mV dec−1
and Cads 60( 10) ␮F cm−2. The change of the transient
electrokinetic behaviour is even more remarkable for aged
electrolyte at pH 6: the adsorption pseudo-capacitance in-
creases to very high values, ranging from 150 to a maxi-
mum of about 500 ␮F cm−2 and the Tafel transient parame-
ter BT becomes lower than 60 mV dec−1, at variance with the
general behaviour summarised in Fig. 12 for not-aged sim-
ple electrolytes. In particular, the lowest BT and the highest
The analysis of the correlation between growth modes
and transient electrokinetic behaviour during deposition can
give interesting insight on Co electrocrystallisation from sul-
phamate solution, permitting the identification of the main
factors determining crystal structure and texture of Co de-
posits. The kinetics–structure correlation is basically deter-
mined by the adsorption behaviour of hydrolysis products,
whose role as reaction intermediates was first recognised by
Simonova and Rotinyan [55] and Heusler [56,57] and, later,
by Lenoir and Wiart [58].
At low pH, Co structure is biphasic fcc+hcp, with possible
predominance of ␤-Co depending on the rate of the hydrogen
discharge reaction. The resulting structure is heavily faulted
as a consequence of accumulation of stacking defects, pos-
sibly twin faults, on fcc {1 1 1} planes. In this growth con-
dition, adsorbed hydrogen is the stable species at the surface
and its presence gives high adsorption pseudo-capacitance
and BT in the range of 2RT/F, which is the typical value for
hydrogen evolution.
The metastable fcc phase slowly vanishes increasing pH
and [11.0] PO comes out, through a transition region charac-
terised by the formation of a faulted hcp structure, with faint
[10.0] and [11.0] texture. This structure is still affected by
the influence of adsorbed hydrogen, as shown by the persis-
tence of BT in the range of 2RT/F for both not-buffered elec-
trolytes (in the pH range 1.8–3) and buffered electrolytes (at
pH 2.7–3.2). Significantly, a similar behaviour is also dis-
played by aged electrolytes (at pH 3.2–3.6), i.e. not-buffered
solution giving coatings with degraded growth structure. The
electrokinetic behaviour is the possible result of the compe-
tition between equally weak adsorption processes and, in-
terestingly, it is only slightly affected by boric acid addition.
The relative predominance of hydrogen or hydrolysed inter-
mediates adsorbed at the interface can explain the step-wise
change of the transient Tafel parameter BT for not-buffered
electrolytes at pH 3.2; whilst for buffered electrolytes, a lin-
ear trend is displayed by BT. The growth structure result-
ing from these surface conditions can be interpreted as the
transition state between the biphasic fcc+hcp and the single
phase hcp structure. In the presence of boric acid, the tran-
sition pH range is slightly shifted towards less acidic value.
More defined structural types are those obtained at pH val-
ues above the transition range: [11.0] and [00.1] PO, with-
out H3BO3; and [10.0] + [11.0] mixed PO structure, with
H3BO3. The corresponding transient kinetic behaviour is
very different.
C
ads value are observed for electrolyte containing sulphamid
5 mM and giving strong [00.1] PO.
In the presence of H3BO3 0.4 or 0.5 M, the transient elec-
trokinetic behaviour is quite different. The SCP parameters
display an opposite variation with increasing solution pH
from 2.7 to 4.8: BT increases almost linearly from about
100 to over 160 mV dec−1; in the same pH range, Cads
decreases from 90 to 60 ␮F cm−2 and this change takes
place over a 0.5 pH interval. In the low pH region of the
above range, SCP parameters show values similar to those
observed in the absence of H3BO3: at pH 2.7, BT and
Cads are: 100 mV dec−1 and 85( 10) ␮F cm−2, respectively;
those derived for not-buffered solution at pH 2.5 and 3 are:
110 mV dec−1 and 65( 10) ␮F cm−2, respectively.
On the other hand, the transient overvoltage at pH >
3.5 is greatly enhanced compared to solutions of same pH
without H3BO3. At pH 4 or higher, SCP parameters are:
BT 160–170 mV dec−1 and Cads from 50 to 70 ␮F cm−2, the
lowest value being observed at the highest pH value; τ is
in the range 5–60 ms at pH 3.8, and 10–20 ms at pH 4.8.
Tafel slopes, resulting from steady-state measurements in
the cd range 10–100 mA cm−2, are 110 mV dec−1 (pH 4.8)
and 125 mV dec−1 (pH 3.8).
Chloride addition to buffered electrolytes induces a sen-
sible increase of both Cads and BT. For Co␴ 1 M, CoCl2
When [11.0] texture develops, outgrowth occurs with low
Cads and BT values as a consequence of stable adsorption of
discharge intermediates. Outgrowth progressively decreases
giving rise to lateral growth as a consequence of electrolyte
ageing, due to the evolving hydrolytic speciation, and with
pH increase. Mono-oriented [00.1] layers are obtained at
high pH, in conditions of strong surface coverage by hydrox-
ides or basic salts and accumulation of reducible species at
2
0.1 M, H3BO3 0.5 M, at pH 5.2, SCP parameters become: BT
180( 10) mV dec−1 and Cads 90( 10) ␮F cm−2. No steady
overvoltage is reached within the short time of the transient
and the t1/2 term must be considered, resulting in τ values
almost constant at about 6 ms. Tafel slopes from both po-
tentiodynamic and steady-state potentiostatic measurements
are 80–90 mV dec−1
.

4088
A. Vicenzo, P.L. Cavallotti / Electrochimica Acta 49 (2004) 4079–4089
the interface, resulting in the highest Cads and the lowest BT
values.
sation of discharge intermediates as complexes stabilised
at the surface by boric acid; the formation of stable nuclei
requires cooperative adsorption and discharge of inter-
mediates, resulting in complex morphological features.
The mixed structure with [10.0] + [11.0] PO is typical
of electrolytes containing boric acid. The electrokinetic
behaviour drastically changes: BT increases up to 3RT/F;
Cads slightly increases with respect to Co baths of same
pH without H3BO3; polarisation effects appear during
the transients. These changes point to different phenom-
ena: the high BT value can be related to adsorption or
chemical contribution to the activation overvoltage; corre-
spondingly, the pseudo-capacitance is relatively high and a
pseudo-diffusional transient overvoltage is observed because
of faradaic and adsorption processes overlapping. Accord-
ing to this interpretation, cobalt discharge takes place from
complex species adsorbed at the surface; these complex
species could be formed by Co(OH)+-borate or polyborates
interaction, since their stability appear to increase as elec-
trolyte acidity decreases. The highest values of both BT and
Cads occur in the presence of CoCl2 0.1 M in the plating
solution, suggesting that chloride are directly involved in
the reaction mechanism stabilising discharge intermediates.
Based on these arguments, the correlation between SCP
transient behaviour and the different growth structure ob-
served in ECD Co from sulphamate solutions can be ra-
tionalised considering the stability and influence upon nu-
cleation and growth of different surface species, whose be-
haviour depends on the electrolyte pH and hydrolytic equi-
libria. Three basic growth modes can be identified:
The nuclei formation, dependent on the substrate or im-
posed by the solution, and the deposits morphology, result-
ing in pyramids or ridges growth, in the absence or presence
of boric acid respectively, point to a specific role of boric
acid in the electrocrystallisation process. In agreement with
the interpretation first proposed by Simonova and Rotinyan
[55] and recently, by Zech and Landolt [59], the kinetic in-
fluence of boric acid could be ascribed to surface interaction
by adsorption or complex species formation.
In conclusion, the kinetics–structure relationship in cobalt
electrocrystallisation from sulphamate electrolyte is demon-
strated by the correlation between the transient Tafel pa-
rameter and growth structure: the highest values of BT in
the range of 3RT/F are found when layer growth proceeds
by independent and continuous nucleation from solutions
with boric acid; lower values are related to different pro-
cesses: (1) hydrogen adsorption and evolution, when BT is
about 2RT/F, resulting in the formation of two-phase or sin-
gle phase highly defective structure; (2) adsorption of stable
hydrolysed species as discharging intermediates, when BT
is slightly higher than RT/F, resulting in self-perpetuating
layer growth; (3) precipitation of hydroxides or basic salts at
grain boundaries, when BT is lower than RT/F, resulting in
cellular crystallisation through lateral spreading of growth
steps.
1. Outgrowth or perpendicular [11.0] textured growth, with
transient Tafel parameter RT/F < BT < 3RT/2F: hydrol-
ysis influence is confined to the interface, where local
stabilisation of intermediate hydrolysed species results
in weak nucleation activity and stable growth; the sur-
face stability of hydrolysed species is reduced by acidity
increase, as shown by both BT and Cads increase. This
is the growth mode characteristic of pure cobalt sulpha-
mate solution at pH > 3.2; it is negatively affected by
solution ageing, degrading towards highly defective and
even amorphous growth structure but can be effectively
stabilised by chloride addition.
References
[1] H. Fischer, Elektrolytische Abscheidung und Elektrokristallisation
von Metallen, Springer Verlag, Berlin, 1954;
H. Fischer, Electrodepos. Surf. Treat. 1 (1972–1973) 239;
H. Fischer, Ibidem 1 (1972–1973) 319.
[2] J. Amblard, M. Froment, N. Spyrellis, Surf. Technol. 5 (1977) 205.
[3] J. Amblard, I. Epelboin, M. Froment, G. Maurin, J. Appl. Elec-
trochem. 9 (1979) 233.
[4] A. Argyriou, N. Spyrellis, Trans. IMF 71 (3) (1993) 83.
[5] H.S. Karayannis, G. Patermarakis, Electrochim. Acta 40 (1995) 1079.
[6] T. Osaka, Electrochim. Acta 45 (2000) 3311.
[7] S. Armyanov, Electrochim. Acta 45 (2000) 3323.
[8] W. Hull, Phys. Rev. 17 (1926) 571.
[9] H. Kersten, Physics 2 (1932) 274.
[10] N.A. Pangarov, St. Rashkov, Comput. Rend. Acad. Bulgare Sci. 13
(1960) 439.
[11] Yu.M. Polukarov, Russ. J. Phys. Chem. 34 (1960) 68.
[12] L. Cadorna, P. Cavallotti, Electrochim. Metall. 1 (1966) 364.
[13] R. Sard, C.D. Schwartz, R. Weil, J. Electrochem. Soc. 113 (1966)
424.
2. Lateral growth, i.e. [00.1] textured growth, with BT
≤
RT/F: hydrolysed species are stable in the bulk solution
and surface coverage is very high; precipitation and in-
corporation of hydrolytic products occurs at the growing
surface without hindering crystal growth; the peculiarity
of these growth conditions leads to cellular crystallisa-
tion. Deposits with strong [00.1] texture are not easily
obtained from not-buffered aged electrolytes at high pH
unless in conditions of epitaxial growth on [1 1 1] tex-
tured Cu substrate; in this case highly [00.1] textured
deposits can be obtained at pH as low as 4.
[14] Y.N. Sadana, Surf. Technol. 5 (1977) 97.
[15] M. Cerisier, K. Attenborough, E. Jedryka, M. Wojcik, S. Nadolski,
C. Van Haesendonck, J.P. Celis, J. Appl. Phys. 89 (2001) 7083.
[16] T.S. Eagleton, J.P.G. Farr, Trans. IMF 80 (2002) 9.
[17] J. Dille, J. Charlier, R. Winand, J. Mater. Sci. 32 (1997) 2637.
[18] T. Cohen-Hyams, W.D. Kaplan, J. Yahalom, Electrochem. Solid-State
Lett. 5 (2002) C75.
3. Cluster growth, with BT (2.5–3) RT/F: this growth mode
corresponds to the [10.0] + [11.0] mixed texture obtained
from H3BO3 buffered solution at pH > 3.5. In the pres-
ence of boric acid, growth kinetics is related to the stabili-
[19] H.L. Gaigher, N.G. Van Der Berg, Electrochim. Acta 21 (1976) 45.

A. Vicenzo, P.L. Cavallotti / Electrochimica Acta 49 (2004) 4079–4089
4089
[20] S. Nakahara, S. Mahajan, J. Electrochem. Soc. 127 (1980) 283.
[21] N.A. Pangarov, S.D. Vitkova, Electrochim. Acta 11 (1966)
1733.
[39] R. Piontelli, G. Poli, G. Serravalle, in: Transactions of the Symposium
on Electrode Processes, E. Yeager (Ed.), Wiley, NY, 1959, p. 67;
R. Piontelli, Electrochim. Metall. 1 (1966) 5.
[22] N.A. Pangarov, J. Electroanal. Chem. 9 (1965) 70.
[23] R. Breckpot, An. R. Soc. Esp. Fis. Quim. 61B (1965) 51.
[24] S. Armyanov, S.D. Vitkova, Surf. Technol. 7 (1978) 319.
[25] G.I. Finch, H. Wilman, L. Yang, Disc. Faraday Soc. 1 (1947) 144.
[26] I.M. Croll, B.A. May, in: Electrodeposition Technology Theory and
Practice, ECS Proc. 87-17, L.T. Romankiw, D.R. Turner (Eds.), ECS
Pub., Pennington, NJ, 1987, p. 295.
[27] M. Froment, G. Maurin, C. R. Acad. Sci. Paris 266C (1968) 1017.
[28] H. Nakano, K. Nakahara, S. Kawano, S. Oue, T. Akiyama, H.
Fukushima, J. Appl. Electrochem. 32 (2002) 43.
[29] J. Scoyer, R. Winand, Surf. Technol. 5 (1977) 169.
[30] A.K.N. Reddy, J. Electroanal. Chem. 6 (1963) 141;
A.K.N. Reddy, S.R. Rajagopalan, Ibidem 36 (1963) 153;
A.K.N. Reddy, S.R. Rajagopalan, Ibidem 6 (1963) 159.
[31] P.L. Cavallotti, E. Galbiati, Tu Chen, in: Electroplating Engineering
and Waste Recycle, D. D. Snyder, U. Landau, R. Sard (Eds.), ECS
Pub., Pennington, NJ, 1983, p.130.
[32] P. Cavallotti, G. Caironi, Surf. Technol. 7 (1978) 1.
[33] Tu Chen, P. Cavallotti, Appl. Phys. Lett. 41 (1982) 205.
[34] Tu Chen, P.L. Cavallotti, IEEE Trans. Magn. 18 (1982) 1125.
[35] P.L. Cavallotti, A. Vicenzo, in: Fundamental Aspects of Electrochem-
ical Deposition and Dissolution, M. Matlosz, D. Landolt, R. Aogaki,
Y. Sato, J.B. Talbot (Eds.), PV 99-33, ECS Pub., Pennington, NJ,
2000, p. 123.
[40] P.L. Cavallotti, D. Colombo, U. Ducati, A. Piotti, in: Electrodepo-
sition Technology Theory and Practice, L. Romankiw, D.A. Turner
(Eds.), PV 87-17, ECS Pub., Pennington, NJ, 1987, p. 429;
P.L. Cavallotti, B. Bozzini, L. Nobili, G. Zangari, Electrochim. Acta
39 (1994) 1123.
[41] A. Vicenzo, P.L. Cavallotti, J. Appl. Electrochem. 32 (2002) 743.
[42] L.O. Olikhovska, Mater. Sci. Forum 321–324 (2000) 103.
[43] B.E. Warren, X-ray Diffraction, Dover Publications, New York,
1990.
[44] O.S. Edwards, H. Lipson, Proc. R. Soc. 180A (1941) 268.
[45] T.H.V. Setty, H. Wilman, Trans. Faraday Soc. 51 (1955) 984.
[46] N.A. Pangarov, V. Velinov, Electrochim. Acta 13 (1968) 1641.
[47] S.K. Verma, H. Wilman, J. Phys. D: Appl. Phys. 4 (1971) 1167.
[48] J. Thevenin, J. Microsc. Spectrosc. Electron. 1 (1976) 7.
[49] C.F. Baes, R.E. Mesmer, in: The Hydrolysis of Cations, Wiley, 1976,
p. 238.
[50] J. Ji, W.C. Cooper, Electrochim. Acta 41 (1996) 1549.
[51] G.K. Williamson, W.H. Hall, Acta Metall. 1 (1953) 22.
[52] I. Epelboin, M. Froment, G. Maurin, Plating 59 (1969) 1356.
[53] J. Bard, L.R. Faulkner, in: Electrochemical Methods, Wiley, New
York, 1980, p. 249.
[54] B.E. Conway, in: Theory and Principles of Electrode Processes, The
Ronald Press Company, New York, 1965, p. 236.
[55] M.V. Simonova, A.L. Rotinyan, Zh. Prikl. Khim. SSSR 37 (1964)
1951.
[36] P.L. Cavallotti, A. Vicenzo, M. Bestetti, S. Franz, Surf. Coat. Technol.
169–170 (2003) 76.
[56] K.E. Heusler, Z. Elektrochem. 66 (1962) 177.
[37] G.B. Harris, Philos. Mag. 43 (1952) 113.
[38] M.H. Mueller, W.P. Chernock, P.A. Beck, AIME Trans. 212 (1958)
39.
[57] K.E. Heusler, Ber. Bunsenges. Phys. Chem 71 (1967) 620.
[58] F. Lenoir, R. Wiart, C.R. Acad. Sci. Paris 269C (1969) 204.
[59] N. Zech, D. Landolt, Electrochim. Acta 45 (2000) 3461.