Electrochemical study of cathodic electroprecipitation of strontium hydroxide films from dimethyl sulfoxide-based solvents

Article abstract of DOI:10.1149/1.1531196

Films of strontium hydroxide and related compounds have been galvanostatically electrodeposited on silver tapes in organic Sr(II) nitrate and chloride solutions using dimethyl sulfoxide (DMSO) because of its high dissolving power and some water as the hydroxide ion precursor. The influence of the nature of the solvent, electrolyte concentrations, and current applied during deposition on film microstructure and composition have been investigated by means of voltammetry, chronoamperometry, and ex situ spectroscopic methods: X-ray diffraction, scanning electron microscopy, electron microprobe analysis, and inductively coupled plasma spectrophotometry. Films were found to be crystalline after formation at ambient temperature whatever the conditions. Optimal current ranges were determined for each electrolyte/solvent couple in terms of adherence and homogeneity of the precursor. The electroprecipitation mechanism has been completed in pure DMSO by a subsequent chemical reaction between strontium hydroxide and atmospheric CO₂ giving rise to crystallized strontium carbonate. In acetonitrile + DMSO mixtures, the chloride phase SrCl₂ · 2H₂O has also been identified as coprecipitated with Sr(OH)₂ because of the low solubility of the electrolyte in pure acetonitrile.

Full text of DOI:10.1149/1.1531196

D30
Journal of The Electrochemical Society, 150 ͑2͒ D30-D40 ͑2003͒
0
013-4651/2003/150͑2͒/D30/11/$7.00 © The Electrochemical Society, Inc.
Electrochemical Study of Cathodic Electroprecipitation of
Strontium Hydroxide Films from Dimethyl Sulfoxide-Based
Solvents
z
C. Lepiller, S. Poissonnet, F. Legendre, and G. Giunchi
CEA/Saclay, Service de Recherches en M e´ tallurgie Physique, 91191 Gif-sur-Yvette, France
Films of strontium hydroxide and related compounds have been galvanostatically electrodeposited on silver tapes in organic Sr͑II͒
nitrate and chloride solutions using dimethyl sulfoxide ͑DMSO͒ because of its high dissolving power and some water as the
hydroxide ion precursor. The influence of the nature of the solvent, electrolyte concentrations, and current applied during depo-
sition on film microstructure and composition have been investigated by means of voltammetry, chronoamperometry, and ex situ
spectroscopic methods: X-ray diffraction, scanning electron microscopy, electron microprobe analysis, and inductively coupled
plasma spectrophotometry. Films were found to be crystalline after formation at ambient temperature whatever the conditions.
Optimal current ranges were determined for each electrolyte/solvent couple in terms of adherence and homogeneity of the
precursor. The electroprecipitation mechanism has been completed in pure DMSO by a subsequent chemical reaction between
strontium hydroxide and atmospheric CO giving rise to crystallized strontium carbonate. In acetonitrile ϩ DMSO mixtures, the
2
chloride phase SrCl •2H O has also been identified as coprecipitated with Sr(OH) because of the low solubility of the electrolyte
2
2
2
in pure acetonitrile.
©
2003 The Electrochemical Society. ͓DOI: 10.1149/1.1531196͔ All rights reserved.
Manuscript submitted June 3, 2002; revised manuscript received July 29, 2002. Available electronically January 2, 2003.
Electrochemistry is a powerful and attractive technique for thin-
film synthesis, following the principles of soft chemistry with its
rational approach based on mechanistic and solution chemistry con-
siderations and kinetic control over the reactions allowing the syn-
thesis of metastable phases at near-ambient temperatures.1
Electrodeposition offers numerous advantages over alternative
deposition methods, including the ability to deposit uniform films on
complex substrate designs, a high degree of purity attainable despite
low production cost, and an easy large-scale adaptation for industrial
implementation. Above all, the control of the driving force for the
reactions involved in deposition by electrochemical parameters re-
sults in a tight control of phase composition, morphology, and depo-
sition rate of the films.
netic and solid fuel cell materials. After formation, electrodeposited
films are usually amorphous and supposed to be composed of hy-
3
4
27,28
35
droxide particles of nanometer size, except for ZnO,
PbO,
3
6
and Cu O, identified as crystallized oxides. Fresh deposits are then
2
converted to corresponding active oxides by a convenient thermal
treatment.
Until today, research activity on electrodeposition of oxides/
hydroxides has mainly focused on demonstrating the feasibility of
the technique for various compounds of technological interest; how-
ever, there is little literature available about the deposition mecha-
nism itself, especially for cathodic processes. There is a general
agreement about the fact that the insoluble precursor is precipitated
at the cathode after reaction between the metal ions and the electro-
Ϫ
Among the large variety of materials accessible to electrodepo-
sition and investigated up to now, metal oxide and hydroxide coat-
ings are of considerable interest owing to their potential practical
generated base OH , as a consequence of the local pH increase. But
the nature and the relative proportions of the electrochemical reac-
tions giving rise to the hydroxide ions remain mostly speculative.
Thermodynamic considerations have led to the proposal that nitrate
reduction to nitrite or ammonium ions must be the main source of
the base in aqueous nitrate solutions, rather than water electrolysis,
2
applications, in continuous expansion. Until recently, transition
metal oxides were the subject of a great number of studies, espe-
cially because of their charge storage capabilities ͑Ni, Mn͒, electro-
chromic activities ͑Ni, Rh, Ir, W͒, electrocatalytic ͑Ru͒, and photo-
3
7
also called the hydrogen evolution reaction ͑HER͒. Following this
3
electrochemical ͑Cu, Ni, Fe͒ properties. As nickel hydroxide is the
assumption, most studies have been undertaken in this electrolyte, as
Ϫ
most commonly used positive electrode in rechargeable batteries,
fabricated by the impregnation technique,4 deposition conditions
from aqueous nitrate solutions have been extensively studied in an
attempt to relate them to the film thickness.5
for Ni(OH) . Nevertheless, the exact origin of OH has not yet
2
been clearly identified. Unlike several groups who were able to dis-
Ϫ
tinguish between H O and NO3 reduction by cyclic
voltammetry
2
-9
30,32
the work of Kamath in Mg(NO3)2 did not reveal
Moreover, ceramic oxides are suited for applications in high-
any specific features assignable with certainty to the nitrate reduc-
10
temperature and corrosive environments. Since the pioneering
38
tion reaction in addition to hydrogen and oxygen evolution. The
11
work by Switzer in 1987, the electrochemical formation of
same group found that the yields of metal hydroxides obtained from
a chloride bath were comparable if not actually higher than that
from a nitrate bath, showing that HER contributes as much to elec-
11-13
14,16
16,17
14,15,18,19
19
CeO2 ,
TiO2 ,
RuO2 ,
ZrO2 ,
and Al O
2
3
was
demonstrated during the last decade, as well as that of composite
coatings1
6,19
and supperlattices. Recently, substantial effort has
20
39
trogeneration of base as the nitrate reduction reaction. In organic
evolved for electronic applications. For instance, the electronic con-
media, the question should seem to be overcome, but it has to be
noted that water is generally present in solution as part of the solvent
mixture or from use of hydrated salts. Thus, the formation of hy-
droxide compounds in organic solvents from nitrate salts has been
ductivity of RuO makes possible its use as a metallic contact
2
electrode21 and its chemical stability makes it an effective diffusion
2
2
23
barrier in silicon-integrated circuits, superconducting thin films,
or ferroelectric devices.24 So a growing interest has resulted from its
electrolytic deposition. In this way, promising results are emerging
4
0,41
noted to occur via the reduction of water,
reduction is still believed the more likely reaction at lower
potentials.
even though nitrate
2
5-28
4
2
with ZnO,
a transparent semiconductor widely used for opto-
electronic applications. Finally, complex oxides such as
Literature is also scarce concerning electrodeposition of unary
hydroxides of alkaline-earth metals, especially in organic solvents.
Strontium is an element of importance in a variety of complex ce-
ramic oxides. Textured high TC Bi-based superconductors have al-
2
9-33
perovskites
can also be prepared successfully by electrodeposi-
tion, enlarging the capabilities of the method further to ferromag-
ready been prepared by electrodeposition, most often in a one-step
z
43-47
E-mail: lepiller@ortolan.cea.fr
route.
Aqueous and nonaqueous electrolytic solutions have been
Downloaded on 2015-06-14 to IP 134.129.120.3 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 150 ͑2͒ D30-D40 ͑2003͒
employed for this purpose; it has been observed that oxides synthe-
D31
sized in water exhibit lower and broader transition temperatures due
to a poor morphology related to simultaneous HER during
electrolysis.48 The best results have been obtained in dimethyl sul-
foxide ͑DMSO͒, because of its high dielectric constant allowing a
quantitative dissolution of a large number of inorganic salts and its
resistance to reduction providing the wide working potential range
needed for alkaline-earth metal deposition.4 In our laboratory, we
are perfecting an original technique for the elaboration of the Bi-
9
50,51
2212 precursor by sequential electrodeposition.
As part of this
process, we aim at getting a better understanding of chemical and
electrochemical mechanisms involved in deposition of Sr-O com-
pounds. In this article, we report the results of an extensive voltam-
metric study of the electrochemical behavior of Sr͑II͒ chloride and
nitrate salts realized in pure DMSO and in organic mixtures contain-
ing DMSO, combined with structural, morphological, and composi-
tional analysis of micrometer-thick films prepared from these differ-
ent solutions.
Experimental
Figure 1. Cathodic current-potential curves in DMSO as a function of the
nature of the Sr͑II͒ salt. Sr͑II͒ concentration: 0.1 M. Scanning rate:
5 mV s . Rotation speed: 500 rpm. Substrate: Pt.
All chemicals were analytical grade reagents from Merck used
without further purification. The solvent consisted of either pure
dimethyl sulfoxide or a ratio of six volumes of acetonitrile to one
volume of DMSO, the latter being added owing to its extensive
dissolving power toward hydrated as well as anhydrous nitrate and
Ϫ1
49
Deposition was carried out at ambient temperature in unstirred
solutions. Due to the insulating character of the deposits, the chosen
halide metallic salts. Organic solvents were Normapur grade and
used as received, taking into account that DMSO is reported to
contain 0.01% residual water. Sr͑II͒ ions were dissolved in the form
of anhydrous SrCl , Sr(NO ) , or sixfold hydrated salts at molar
mode was galvanostatic. Current densities ranged from 0.25 to
Ϫ2
2
0 mA cm , for deposition times varying between 30 s and 100
2
3 2
min. After deposition, the films were pressed between two Kim-
wipes and then dried in air at 60°C for 30 min. Some of them were
annealed in air at 880°C for 5 min.
Elemental compositions were determined using an ICP Modula
spectrophotometer from Spectroflame Analytical Instruments. The
layers were dissolved in 15% nitric acid by volume and then pre-
cisely diluted to 100 mL with distilled water prior to analysis.
The morphology and the microstructure of the layers were exam-
ined by scanning electron microscopy ͑SEM͒ with a FEG Gemini
concentrations of 0.1 or 0.5 M unless otherwise stated. No conduc-
tive supporting electrolyte was needed under these conditions. In
solvent mixtures, Sr(OH) •8H O was introduced in excess in order
2
2
to neutralize in situ the protons liberated by the anodic reaction and
to buffer the whole solution. This specific point is discussed in the
next section. All solutions were deaerated with argon gas bubbling
prior to each electrodeposition and flow was continued over the
solution during measurements.
Electrochemical experiments were carried out with
a
15-25 from LEO at an emission voltage of 5 keV, while quantitative
potentiostat-galvanostat Autolab model PGSTAT 100 from Ecoche-
mie, interfaced to a PC equipped with the GPES 4.5 software. Vol-
tammetry investigations were conducted in a Metrohm undivided
cell setup ͑volume 50 mL͒ in the standard three-electrode configu-
ration. For fundamental studies under controlled hydrodynamic con-
ditions, a platinum rotating disk electrode ͑diameter 2 mm͒, Voltalab
model CTV 101 from Radiometer, served as the working electrode
in conjunction with a platinum wire as the auxiliary electrode and a
saturated calomel electrode ͑SCE͒ as the reference. The choice of
the SCE lies in the fact that the liquid junction potential between
atomic analysis and X-ray cartographies were performed using a
Cameca SX 50 electron microprobe. Finally, the crystalline structure
and phase composition were determined by the X-ray diffraction
͑XRD͒ method. The diffraction patterns were recorded with a Phil-
ips X-ray generator model PW 1820 using monochromatized Cu K␣
radiation and operating at 40 kV and 28 mA ͑scanning speed
Ϫ1
0
.6° min ).
Results and Discussion
Electrochemical study.—Figure 1 displays the j-E curves ob-
tained under hydrodynamic conditions from 0.1 M strontium chlo-
ride and 0.1 M strontium nitrate in DMSO at a platinum rotating
electrode. In each case a single reduction wave is observed, com-
mencing at Ϫ1.6 V in nitrate solution and at Ϫ1.8 V in chloride,
aqueous potassium chloride and DMSO electrolytes is known to be
stable within Ϯ3 mV.52 The liquid junction potential was thus esti-
mated to be approximately 120 mV by equating the redox potentials
of large ions in the two media, which we have confirmed on our own
with the Cu2ϩ/Cu couple ͓⌬E ϭ 116 mV in 5 mM Cu͑II͔͒. The
reference electrode was separated from the solution by a fritted
characterized by a half-wave potential of Ϫ2 V. Limiting currents
bridge containing 0.1 M KNO to prevent contamination. All poten-
3
Ϫ2
starting at Ϫ2.5 V are of comparable intensity around 4 mA cm
,
tials hereafter are quoted with respect to the SCE.
The electrochemical cell used for deposition ͑volume 250 mL͒
included the cathodic substrate centered between two parallel plati-
num grids as counter electrodes. The substrates were silver tapes
thickness 50 ␮m; electroactive area 14 cm ) suspended vertically
though slightly higher ͑in absolute value͒ in the nitrate solution. The
Tafel approximation is applicable at low overpotentials where the
cathodic current is under charge-transfer control and the reverse
2
oxidation reaction may be neglected. Tafel plots log͉j͉-E obtained
͑
from the j-E curves of Fig. 1 in the charge-transfer region (Ϫ1.95
Ͻ E Ͻ Ϫ1.75 V) are reported on Fig. 2. The slope of the linear
part, ␣nF/2.3RT, gives comparable values of ␣, 0.23 and 0.17, for
chloride and nitrate solutions respectively ͑with n ϭ 2, see below͒.
Such small charge-transfer coefficients are indicative of the irrevers-
ibility of the reduction process, as most often found in DMSO.49
Taking into account that the order of the half-wave potentials in
and fixed to the stainless steel sample holder by means of a locking
screw making electrical contact. This allowed one to maintain a
reproducible geometric configuration in the deposition cell.
The silver slides were prepared as follows: once cut, they were
first degreased by dipping in Viapred SID cleaning solvent, then
rinsed with distilled water, alcohol, and acetone consecutively. After
drying in air and just before deposition, the surface was thermally
activated for 10 min at 880°C in air, causing recrystallization of
large silver grains along the preferential orientation ͗200͘.
5
2
DMSO is similar to that in aqueous solutions, three successive
reactions are predictable from thermodynamic data
Downloaded on 2015-06-14 to IP 134.129.120.3 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D32
Journal of The Electrochemical Society, 150 ͑2͒ D30-D40 ͑2003͒
Figure 3. Cathodic current-potential curves in DMSO as a function of water
Ϫ1
concentration. Sr͑II͒ concentration: 0.1 M. Scanning rate: 5 mV s . Rota-
tion speed: 500 rpm. Substrate: Pt.
Figure 2. Tafel plots, log͉j͉ ϭ f(E), in the charge-transfer region as a func-
Ϫ1
tion of the Sr͑II͒ salt. Sr͑II͒ concentration: 0.1 M. Scanning rate: 5 mV s
.
Rotation speed: 500 rpm. Substrate: Pt.
Electrogeneration of base causes a local pH rise in the vicinity of
the cathode, leading to the consecutive chemical precipitation of the
Ϫ
Ϫ
°
O ϩ 2H O ϩ 4e → 4 OH
E ϭ 0.401 V/NHE
͓1͔
2
2
1
2
ϩ
insoluble strontium hydroxide between Sr
ions present in the
Ϫ
Ϫ
Ϫ
Ϫ
°
NO3 ϩ H O ϩ 2e → NO2 ϩ 2 OH
E ϭ 0.01 V/NHE
2
Nernst layer and hydroxide ions diffusing away from the cathode
2
͓2͔
Sr2 ϩ 2 OH → Sr͑OH͒2
ϩ
Ϫ
͓6͔
Ϫ
Ϫ
°
3
2
H O ϩ 2e → H ϩ 2 OH
E ϭ Ϫ0.828 V/NHE ͓3͔
2
2
On the basis of this mechanism of electroprecipitation, Sr2 ions are
ϩ
Sr2 reduction into elemental Sr has been experimentally evidenced
ϩ
Ϫ
not reduced but only undergo chemical reaction with OH . Sr(OH)
49
2
at Ϫ2.1 V vs. SCE for Sr͑II͒ nitrate salts in dry DMSO, but this
reaction is not supposed to occur here because traces of residual
water, more easily reducible than Sr2 , are still present in the sol-
vent and prevent the formation of a metallic deposit, which is in turn
unstable in air and would be readily oxidized if formed.
In dimethyl sulfoxide, due to the strong stabilization of the su-
peroxide anion, oxygen reduction is actually separated into two suc-
cessive one-electron steps characterized by half-wave potentials of
Ϫ0.85 and Ϫ2.2 V for alkali metal and quaternary ammonium
34
is deposited in the form of nanometric colloidal particles.
Figure 3 shows the cathodic polarization curve obtained in a
sixfold hydrated strontium chloride salt as compared to that previ-
ϩ
ously obtained in anhydrous SrCl . According to Nernst’s law, the
2
increase in water contentration leads to a positive shift of the reduc-
tion wave onset from Ϫ1.8 to Ϫ1.5 V and to an increase of its
Ϫ2
height from 4 to 11 mA cm . However, the limiting current is not
proportional to the H O content if we assume an initial value of 6
2
Ϫ3
5
3
ϫ 10 M in pure DMSO ͑corresponding to 0.01% by volume͒,
salts
and is insensitive to hydrodynamic conditions. This is due to the
interference by the chemical step of formation of an insulating hy-
droxide layer depositing at the cathode simultaneously with water
electrolysis. The process is surface controlled and not diffusion
controlled.
The electrochemical behavior of strontium salts plus water has
been further examined on silver tape electrodes under static condi-
tions. On Fig. 4 are superimposed the cathodic polarization curves
obtained up to Ϫ10 V in hydrated 0.1 M strontium nitrate, 0.1 and
0.5 M strontium chloride, respectively. In good agreement with our
Ϫ
Ϫ
O ϩ e → O2
͓4͔
͓5͔
2
Ϫ
Ϫ
O2 ϩ e ϩ cation → peroxide
As our experiments were performed in an inert gas atmosphere,
oxygen reduction need not be considered. However, the first step
was found at Ϫ0.8 V in nondeaerated 0.1 M KNO , under diffu-
3
sional control, which is consistent with the cited literature.
The reduction wave taking place at Ϫ2 V in both chloride and
nitrate medium is therefore assumed to be mainly due to water elec-
trolysis ͑Reaction 3͒. Indeed, it is associated with the formation of a
solid deposit and to the simultaneous liberation of gas bubbles at the
surface of the cathode. As in a previous work performed in aqueous
observations made on platinum electrodes, for the same concentra-
Ϫ2
tion the limiting current is higher in nitrate medium (11 mA cm
)
Ϫ2
than in chloride medium (8 mA cm ), with the effect being even
more pronounced. Increasing the electrolyte concentration by a fac-
tor of 5 results in an increase of the limiting current only by a factor
of 3. In all cases, the shallow slope of the reduction wave reveals the
strong irreversible character of the process, while the lower current
densities observed on Ag are indicative either of a less catalytic
38
Mg(NO ) , no specific feature of nitrate reduction was visible in
3
2
our voltammograms, except that the limiting current was somewhat
higher in nitrate medium (ϩ10%). This can be explained either by
the intervention of nitrate reduction in the same potential range as
hydrogen evolution, or by a complexing effect of Cl ions toward
Sr , inhibiting their diffusion to some extent. Conversely, in nitrate
medium Sr ions are supposed to be fully uncomplexed and to
diffuse more rapidly. Nevertheless, the fact that the very high nitrate
ion concentration ͑0.2 M͒ causes an almost negligible increase in
reduction current argues against nitrate reduction. So we assume the
complexation of Sr ions to be mainly responsible for the discrep-
ancies observed.
Ϫ
surface toward the H evolution reaction or a higher covering rate of
2
2
ϩ
the insulating strontium hydroxide deposit.
2
ϩ
Figure 5 shows the influence of the nature of the solvent on the
j-E curves obtained in 0.1 M SrCl •6H O under hydrodynamic
2
2
conditions on a platinum electrode, and Fig. 6 displays those ob-
tained under static conditions on Ag. The electrochemical features
relative to water reduction appear quite different in pure DMSO as
opposed to the acetonitrile ϩ DMSO mixture. On the one hand, as
2
ϩ
Downloaded on 2015-06-14 to IP 134.129.120.3 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 150 ͑2͒ D30-D40 ͑2003͒
D33
Figure 4. Cathodic current-potential curves in pure DMSO as a function of
Figure 6. Cathodic current-potential curves in SrCl •6H O as a function of
the nature and concentration of the electrolyte. Scanning rate: 10 mV sϪ1
No stirring. Substrate: Ag.
.
2
2
the nature of the solvent. Sr͑II͒ concentration: 0.1 M. Scanning rate:
Ϫ1
10 mV s . No stirring. Substrate: Ag.
previously described a simple reduction wave is observed in DMSO,
commencing at Ϫ1.5 V and followed by a flat plateau of
Ϫ2
1
1 mA cm . On the other hand, in the solvent composed of
wave is partly due to the liquid junction potentials between aqueous
saturated KCl contained in the reference compartment and the dif-
ferent solvents. Using in first approximation the redox system
acetonitrile ϩ DMSO, the reduction wave appears much less in-
tense (2 mA cm ), preceded by a prewave and a cathodic peak
centered at Ϫ1.5 V. This sharp peak is the signature of an additional
surface step in the reduction mechanism. Adsorption of reaction
products at the electrode surface is known to lower the charge-
Ϫ2
2
ϩ
Cu /Cu, the junction potential, ⌬E(X → H2O), has been evalu-
ated to be 120 mV for X ϭ DMSO and 230 mV for X
ϭ acetonitrile ϩ DMSO ͑6:1͒. The cathodic prepeak distinct from
the reduction wave is visible but less pronounced in the mixture as
at a platinum electrode, and some additional current irregularities
are present in the j-E curve relative to the mixed solvent. The gen-
eral trend in the mixed solvent is still a decrease of the current at
high overpotentials, whereas in pure DMSO solution j remains con-
stant with a flat plateau extending up to Ϫ10 V. At this stage it can
5
4
transfer energy, giving rise to prepeaks in voltammetric curves.
The one observed in the present study may therefore be ascribed to
adsorption of hydroxide ions produced by water reduction at the
platinum cathode. The occurrence of this reaction is expressed both
by a 500 mV positive shift of onset of the cathodic current and by a
strong inhibiting effect in the limiting region. Indeed, the current
tends to decrease after adsorption instead of developing into the
large reduction wave characteristic of DMSO solutions. This behav-
ior could be the consequence of competition between adsorption of
hydroxide ions and their chemical precipitation with Sr2 when re-
leased from the cathode.
thus be argued that H O reduction and concomitant Sr(OH)2 pre-
2
cipitation are somewhat complicated by surface side reactions re-
lated to the chemical solubility of the reactive species. This point is
elucidated in the next part.
ϩ
This experimental feature is much less apparent at a silver cath-
ode, as shown in Fig. 6. First, the potential shift of the reduction
Deposition experiments.—In DMSO solution.—Galvanostatic
deposition conditions have been determined by interpolation from
the polarization curves so that the applied currents would be mark-
edly lower than the limiting value, in the charge-transfer controlled
potential range. This is a key element for electrodeposition of
smooth films, whereas a dendritic growth is expected under higher
11
current conditions. In Table I are reported deposition conditions
nature of the electrolyte, concentration, current density, charge͒ and
͑
masses measured after drying, while Fig. 7a and b display the re-
spective variations of the cathode potential with electrical charge in
0.1 M anhydrous and sixfold hydrated strontium chloride salts. Ac-
cording to the electrochemical study, the cathode potential during
deposition is more negative when the applied current density is
higher. After an initial fall during the first few seconds due to
double-layer charging, the overall evolution consists of a potential
decrease in relation to the formation of the insulating strontium hy-
droxide layer which forces the system toward higher overpotentials
as it grows.
In hydrated SrCl ͑Fig. 7b͒ the potential decreases monotonously
2
with charge whatever the current density, more markedly for high
than for low j values. Such a continuous variation is indicative of a
single reaction mechanism during deposition. Potentials ranging
from Ϫ2 to Ϫ3.4 V are located in the reduction wave, as compared
with the j-E curve of Fig. 4. The potential range is comparable to
the deposition series performed in hydrated Sr(NO ) , indicating
Figure 5. Cathodic current-potential curves in SrCl •6H O as a function of
2
2
the nature of the solvent. Sr͑II͒ concentration: 0.1 M. Scanning rate:
3
2
Ϫ1
5
mV s . Rotation speed: 500 rpm. Substrate: Pt.
that the process involved is the same in both cases.
Downloaded on 2015-06-14 to IP 134.129.120.3 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D34
Journal of The Electrochemical Society, 150 ͑2͒ D30-D40 ͑2003͒
Table I. Conditions for Sr„OH… deposition in pure DMSO solutions, as a function of electrolyte nature and concentration. No stirring.
2
Substrate: Ag.
j (mA cmϪ2
)
Q1 (C)
10Ϫ3
g
Q2 (C)
Mi 10Ϫ3
-
g
Electrolyte
SrCl2
Concentration ͑M͒
0.1
0.25
16.8
10.6
11.8
13.1
13.0
11.3
10.6
11.3
12.0
12.5
13.4
4.0
-
0
1
2
4
.5
SrCl •6H O
0.1
0.5
0.25
16.8
8.1
28
18.3
18.7
19.9
2
2
0
1
2
4
.5
SrCl •6H O
0.25
16.8
10.0
10.3
11.4
11.0
13.5
13.0
13.6
13.4
13.7
13.7
-
2
2
0
1
2
5
7
0
2.8
5
0
.5
4.54
5.62
6.06
6.36
6.48
6.47
6.0
6.18
5.85
9.63
9.20
9.32
9.66
9.56
9.22
9.31
9.94
9.96
10.48
1
1
1
2
Sr(NO ) •6H O
0.5
0.25
16.8
-
3
2
2
0
1
2
5
7
0
2.8
5
0
.5
1
1
1
2
In anhydrous SrCl2 ͑Fig. 7a͒ the evolution of the potential is
quite similar for low j values, involving a continuous negative shift
in the range Ϫ2.2 to Ϫ3 V. By contrast, for current densities of 2
but thinner over the substrate grain boundaries, as detected by the
electron microprobe with a weak Ag signal. Table II displays the
average atomic ratios O/Sr measured for the series of films depos-
ited from 0.5 M strontium nitrate and chloride solutions. The quan-
titative analysis reveals a good homogeneity of repartition of Sr and
O elements, especially for layers grown in chloride medium with
Q ϭ 16.8 C. Fresh deposits exhibit an atomic ratio O/Sr comprised
between 2 and 3, regardless of the current density but depending on
both the charge passed and the nature of the counter ion, as is dis-
cussed further below.
Ϫ2
and 4 mA cm , the E-Q curves exhibit potential oscillations to-
gether with the overall decay, whose amplitude is the largest at
Ϫ2
4
mA cm . From our observations of the cathode surface during
deposition, the variation of the potential is strongly correlated to the
size of cogenerated hydrogen bubbles. Small bubbles are actually
associated with a potential decrease while big ones are formed dur-
ing a potential increase. Hence, oscillations may arise from a peri-
odic lack of water molecules available for reduction in the vicinity
of the cathode. This lack is supposed to be compensated by a sudden
increase of the overpotential in order to maintain the imposed cur-
rent, before a return to regular deposition conditions. In the case of
a monotonous variation of E, hydrogen bubbles are hardly visible
and are expelled from the deposit surface as soon as they form.
An important point is the great stability of the electrolytic DMSO
bath towards acidification due to anodic water oxidation. It is char-
acterized throughout the deposition procedures by a constant pH
value in the range 3.5-4. This means that the main anodic reaction,
unidentified yet, is not liberating protons in the solution as does
water oxidation. Polymer formation or oxidation to dimethylsulfone
are strongly suspected.55
Films formed in hydrated 0.5 M nitrate solution are adherent but
less homogeneous and cracked in the tens of micrometers scale ͑Fig.
9͒. Fresh deposits are formed of rounded fibrous and cauliflower-
like grains in the range of 1 to 5 ␮m in diameter closely packed onto
the surface, with cracks originating from drying stresses during sol-
vent evaporation. No unequivocal morphological evolution was ob-
served with current density.
Figure 10a shows the XRD pattern of an as-grown film deposited
Ϫ2
at 1 mA cm in hydrated 0.1 M Sr͑II͒ chloride, while Fig. 10b
displays the XRD pattern obtained for a typical film grown from
Sr͑II͒ nitrate. In both cases the films are crystallized since formation.
Films deposited from chloride solutions are not composed of a mix-
ture of orthorhombic Sr(OH)2 and Sr(OH) •H O as observed for
2
2
From a morphological point of view, the as-grown layers appear
more dense, adherent, and less cracked as the deposition current is
those deposited from nitrate solutions ͑Fig. 10b͒, but the predomi-
nant phase is SrCO ͑Fig. 10a͒. The main diffraction peaks corre-
3
Ϫ2
decreased from 4 to 1 mA cm in 0.1 M Sr͑II͒ DMSO solution and
spond to the powder diagram of orthorhombic strontium carbonate,
without any preferential orientation. Sr(OH)2 if present appears in
its eightfold hydrated tetragonal form. Intercalation of carbonate
ions in nickel hydroxide layers has already been proved by infrared
Ϫ2
from 20 to 5 mA cm in 0.5 M solution. Figure 8a shows the
smooth needle-like surface representative of the series prepared in
chloride medium. The layers are compact, usually formed of misori-
ented needles under low current conditions and more filamentous at
higher j values. Annealed layers are dense, homogeneous, and take
on the silver underlayer grain boundaries, as seen in Fig. 8b. Cracks
present after formation have disappeared and spherical grains a few
micrometers in diameter have joined, covering the surface entirely
spectroscopy for materials deposited by electrochemical impregna-
tion of porous nickel plates in the 20-60°C temperature range.3
9,56
In
this case, carbonate ions partially exchange from other intercalated
ions ͑chloride or nitrate͒ in order to compensate for hydroxide va-
cancies. In our study however, carbonates are present as the crystal-
Downloaded on 2015-06-14 to IP 134.129.120.3 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 150 ͑2͒ D30-D40 ͑2003͒
D35
Figure 8. Surface micrographs of films deposited in 0.1 M SrCl •6H O.
2
2
Ϫ2
Solvent: pure DMSO. Substrate: Ag. j ϭ 1 mA cm . Q ϭ 28 C. ͑a, top͒
As-grown deposit, ͑b, bottom͒ sintered deposit.
͑Table II͒. This result is consistent with only partial transformation
of Sr(OH) . The proportion of strontium hydroxide in the crystal-
2
Figure 7. ͑a, top͒ Variation of the cathode potential as a function of the
line phase, in a more or less hydrated form, is found to increase with
the thickness of the layer, according to the fact that formation of
carbonate may occur at the cathode surface and then expand into the
layer.
deposition time for films deposited in 0.1 M SrCl . Solvent: pure DMSO.
2
No stirring. Substrate: Ag. ͑b, bottom͒ Variation of the cathode potential as a
function of the deposition time for films deposited in 0.1 M SrCl •6H O.
2
2
Solvent: pure DMSO. No stirring. Substrate: Ag.
Figure 10b shows that films deposited from nitrate DMSO solu-
tions exhibit a much lower SrCO -phase content, whatever j. Dif-
3
fraction patterns give evidence that the layers are mainly composed
of anhydrous and onefold hydrated orthorhombic strontium hydrox-
line compound SrCO , which supports the hypothesis of a chemical
3
reaction with carbon dioxide due to the great affinity of hydroxides
for CO2
Sr͑OH͒2(s) ϩ CO2(sol) → SrCO3(s) ϩ H O
͓7͔
Table II. Quantitative analysis of films deposited in pure
DMSO from Sr„II… nitrate and chloride salts, as determined by
the electron microprobe technique, with different Q values.
SrCl "6H O and Sr„NO … "6H O concentrations: 0.5 M. No
2
(l)
Owing to the intrinsic stability of strontium carbonate, this reaction
is allowed to occur at 298 K. It is characterized by a free enthalpy of
2
2
3
2
2
Ϫ1
stirring. Substrate: Ag.
Ϫ101.1 kJ mol calculated from thermodynamic data.
Compositions given in Table II indicate that the atomic ratio
O/Sr relative to the first series in chloride solution, with Q
ϭ 8.1 C, is comprised between 2.5 and 3 without exceeding this
value within the uncertainty of the method, whereas this ratio is
significatively lower for the second series with Q ϭ 16.8 C, in the
range 2.2-2.6. In both cases no significative evolution is observed as
a function of the deposition current density, what confirms a similar
composition for all films synthesized in the same series. Films char-
acterized by Q ϭ 8.1 C, which are thin enough to undergo an al-
most complete transformation of the hydroxides into carbonates,
exhibit a ratio O/Sr tending to 3 correlated to the main crystalline
phase SrCO present together with some unreacted Sr(OH) . On
Chloride solution Chloride solution Nitrate solution
Q ϭ 8.1 C
Q ϭ 16.8 C
Q ϭ 16.8 C
j (mA cmϪ2
)
Atomic composition O/Sr
0
0
.25
.5
-
2.41 Ϯ 0.09
2.43 Ϯ 0.08
2.54 Ϯ 0.16
2.19 Ϯ 0.17
2.36 Ϯ 0.08
2.45 Ϯ 0.11
2.23 Ϯ 0.11
2.63 Ϯ 0.13
2.54 Ϯ 0.23
2.57 Ϯ 0.15
2.25 Ϯ 0.26
2.51 Ϯ 0.42
2.36 Ϯ 0.38
2.58 Ϯ 0.24
2.64 Ϯ 0.47
3.04 Ϯ 0.35
2.72 Ϯ 0.24
2.78 Ϯ 0.21
2.94 Ϯ 0.47
2.88 Ϯ 0.31
2.96 Ϯ 0.32
2.80 Ϯ 0.19
3.02 Ϯ 0.29
.77 Ϯ 0.23
.55 Ϯ 0.29
.72 Ϯ 0.18
3.09 Ϯ 0.30
2.67 Ϯ 0.11
2.92 Ϯ 0.32
1
2
5
7
10
12.8
15
2
2
2
3
2
the other hand, thicker deposits (Q ϭ 16.8 C) have an overall lower
oxygen content, but are characterized by O/Sr ratios greater than 2
20
Downloaded on 2015-06-14 to IP 134.129.120.3 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D36
Journal of The Electrochemical Society, 150 ͑2͒ D30-D40 ͑2003͒
Figure 9. Surface micrographs of an as-grown film deposited in 0.5 M
Ϫ2
Sr(NO ) •6H O. Solvent: pure DMSO. Substrate: Ag. j ϭ 5 mA cm
.
3
2
2
Q ϭ 16.8 C.
ide. The presence of carbonates is attested over 2 mA cmϪ2 and
increases with the deposition current density up to values estimated
from relative diffracted intensities around 20% for
j
Ϫ2
ϭ 20 mA cm . In any case carbonate formation is very scarce in
nitrate medium. This reaction seems therefore sensitive to the solu-
tion composition, more present in chloride medium as compared to
nitrate. This means that it does not occur in air once the films are put
into the atmosphere, but more likely in solution where the reactivity
of dissolved carbon dioxide is expected to depend on solution
chemical parameters.
As the relative peak intensities of SrCO ͑111͒ at 25.2° and ͑201͒
3
Sr(OH) •8H O at 32.4°, measured for films grown in chloride so-
2
2
lution, have been found to be almost identical whatever the deposi-
tion current density with a ratio close to 10, we have considered a
rough composition of 90% SrCO ϩ 10% Sr(OH) •8H O in this
Figure 10. ͑a, top͒ XRD pattern of an as-grown film deposited in 0.1 M
3
2
2
Ϫ1
Ϫ2
SrCl •6H O. Solvent: pure DMSO. Substrate: Ag. j ϭ 1 mA cm . Q
case and hence an approximate molar mass of 159.4 g mol . As
cathodic electrolytic deposition is governed by Faraday’s law, we
were able to determine the masses expected with a deposition yield
of 100% from the charge passed, by assuming a biphasic composi-
tion and a two-electron transfer for each Sr deposited, as predicted
by Reaction 3. The ratio of experimental to theoretical masses, rela-
tive to the deposition series performed in 0.5 M hydrated strontium
chloride DMSO solution, was thus reported as a function of the
current density in Fig. 11. A second independent method was used to
determine the faradaic efficiency in chloride and nitrate solutions; in
both cases the amount of strontium was determined by inductively
coupled plasma ͑ICP͒ analysis and reported against that deduced
from the electrical charge. As a set of comparison, chemical quan-
titative determination of the Sr content was performed in parallel
with the deposition series relative to 0.5 M hydrated strontium ni-
trate DMSO solution.
2
2
ϭ 16.8 C. ͑b, bottom͒ XRD pattern of an as-grown film deposited in 0.5 M
Sr(NO ) •6H O.
Solvent:
ϭ 10 mA cm . Q ϭ 16.8 C.
pure
DMSO.
Substrate:
Ag.
j
3
2
2
Ϫ2
Figure 11 shows the strong correlation between the two tech-
niques within the uncertainty of the measurements and assumptions
made; chemical analysis is found to be in excellent agreement with
values deduced from experimental masses assuming a 10%
Sr(OH) •8H O content. The deposition yield, which reflects the
2
2
ratio of electrogenerated hydroxide ions being precipitated with
strontium ions at the cathode surface, increases in chloride solution
Ϫ2
with current density from 70 to 95% until 5 mA cm and then
Ϫ2
reaches a plateau; for j values over 5 mA cm the average yield is
constant around 95%, always larger than 90%, irrespective of the
thickness of the deposit ͑expressed by the charge͒. Conversely, the
deposition yield observed in nitrate solution continuously decreases
with j from 85% between 0.25 and 10 mA cm before stabilizing
at 70-75%. This is far below the values attained in chloride solution
and confirms previous results reported by Annal Therese and
Figure 11. Variation of the deposition yield in 0.5 M SrCl •6H O and in 0.5
2
2
Ϫ2
M Sr(NO ) •6H O as a function of the current density, as determined by
3
2
2
ICP chemical analysis. Comparison with mass measurements. Solvent: pure
DMSO. Substrate: Ag.
Downloaded on 2015-06-14 to IP 134.129.120.3 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 150 ͑2͒ D30-D40 ͑2003͒
D37
III lists deposition conditions and experimental masses. The study
was divided into two sets of experiments, the first series was per-
formed without any additive and the second series with strontium
hydroxide introduced in excess, then filtered. This experimental fea-
ture is explained below.
During the preparation of the first set of samples in acetonitrile-
based solution, we noticed progressive yellowing and acidification
accompanied by a mass decrease of the deposits ͑data of Table III
are given in the chronological order of experiments͒. The pH fell
from an initial value of 10 to a value of 2.5 after five deposition
procedures. We interpret this as the consequence of the oxidation of
water at the platinum anode, releasing protons into the solution dur-
ing deposition in the following pathway
ϩ
Ϫ
2
H O ϭ O ϩ 4H ϩ 4e
͓8͔
2
2
Acidification is deleterious toward precipitation of strontium hy-
droxide because it consumes hydroxide ions formed at the cathode
and promotes the formation of soluble complexes. Consequently,
water oxidation leads to a rapid decrease of the faradaic efficiency
which is responsible for the mass loss observed. We assessed this
hypothesis by calculating the expected pH from Faraday’s law with
one proton produced per electron transferred ͑see Eq. 8͒ and ob-
tained a pH of 2.43 very close to the experimental value of 2.5. The
good correlation between experimental and calculated pHs gives us
confidence that the decay in yield definitely originates from water
oxidation. To minimize this effect, a buffering agent was added in
order to neutralize protons in situ. This was achieved with
Sr(OH) •8H O, introduced in excess in the bottom of the electro-
Figure 12. Variation of the cathode potential as a function of the deposition
time for films deposited in 0.1 M SrCl •6H O. Solvent: acetonitrile
2
2
ϩ DMSO ͑6:1 vol͒. No stirring. Substrate: Ag. Superimposed E-t curves
are relative to Sr(OH) •8H O-buffered solution.
2
2
Kamath,39 who noted higher yields of metal hydroxide deposition in
chloride baths as compared to nitrate ones. Such a feature is contra-
dictory with the general assertion that reduction of nitrate ions into
nitrites and/or ammonia is the main source of base in hydroxide
electroprecipitation. It seems that nitrate reduction, if present simul-
taneously with hydrogen evolution in nitrate solution, might be fol-
lowed by a chemical reaction leading to soluble unidentified species
2
2
lytic cell. This compound, insoluble in the solvent and inactive to-
ward deposition, dissolves progressively by Reaction 9 with protons
Ϫ
2ϩ
and consuming a non-negligible part of the electrogenerated OH
as they are formed, liberating water and Sr ions
ions, therefore lost for hydroxide formation. Anyway, our results
indicate that the faradaic efficiency for Sr(OH) electroprecipitation
ϩ
2ϩ
2
Sr͑OH͒ •8H O ϩ 2H → Sr ϩ 10H O
͓9͔
2
2
2
from chloride solutions is close to 100% provided that the overpo-
tential during deposition is sufficiently high to maintain an appro-
Ϫ
priate rate of hydroxide formation, v. This means that the OH and
This dissolution reaction effectively prevents the acidification of the
solution. Nevertheless, filtration of the solution was not satisfactory
for a durable stabilization of the pH since the masses of deposits
started to fall after four experiments ͑see Table III͒. In consequence
Sr(OH)2•8H2O has to be kept in excess in the bath. In this way, the
pH is stabilized in the range 4.1 Ϯ 0.2, and no significative aging of
the bath was noted after even ten deposition procedures. An alterna-
tive solution would have been to separate anodic and cathodic com-
partments in the cell, but ours was not designed to readily achieve
this purpose. So we solved the problem by adding a buffer inside the
bath retaining the same cell geometry.
On voltammetric curves, the addition of a buffering agent into
the acetonitrile-based solvent leads reproducibly to a decrease of the
cathodic current due to water reduction, as shown in Fig. 13. Stron-
tium hydroxide, highly insoluble in organic solution, is thus sup-
posed to coordinate water molecules preferentially, making them
2
ϩ
Sr concentrations have values such that the solubility product of
Sr(OH)2 is exceeded in the vicinity of the electrode. As the stron-
tium concentration is constant and independent of electrochemical
parameters, formation of hydroxide ions, controlled by the current
density, is the determining factor for the process. In our experimen-
tal conditions this requirement seems to be achieved for j
Ϫ2
у 5 mA cm in 0.5 M Sr͑II͒ solutions. Below this value, the rate
v, which is exponentially related to deposition voltage E by the
Butler-Volmer equation, is too low for a significant rise in interfacial
4
2
pH and subsequent precipitation of strontium hydroxide. There-
fore, a non-negligible part of hydroxide ions diffuse away from the
cathode and do not contribute to the formation of the deposit.
In acetonitrile ϩ DMSO (6:1) solution.—E-t curves relative to
acetonitrile ϩ DMSO solution are reported in Fig. 12, while Table
Table III. Conditions for Sr„OH… deposition in acetonitrile-based solvent mixture. SrCl "6H O concentration: 0.1 M. No stirring. Substrate:
2
2
2
Ag.
j (mA cmϪ2
)
Q ͑C m͒
10Ϫ3
g
Solvent
Electrolyte
SrCl •6H O
Concentration ͑M͒
4
2
1
0.5
0.25
4
2
1
28
28
25.2
28
28
16.8
16.8
16.8
11.7
16.8
24.9
21.1
15.2
11.6
9.8
10.8
12.8
13.1
8.0
Acetonitrile
ϩ
DMSO
0.1
2
2
͑6:1͒
Acetonitrile
ϩ
DMSO
SrCl •6H O
0.1
2
2
ϩ
Sr(OH) •8H O
Saturated
Filtered
0.5
0.25
2
2
͑6:1͒
7.6
Downloaded on 2015-06-14 to IP 134.129.120.3 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D38
Journal of The Electrochemical Society, 150 ͑2͒ D30-D40 ͑2003͒
Figure 13. Influence of Sr(OH) •8H O addition in the electrolytic bath on
2
2
current-potential curves. Solvent: acetonitrile ϩ DMSO. Substrate: Pt. Scan-
Ϫ1
ning rate: 20 mV s . 5th cycling scan. No stirring.
less available for reduction. But it may have no detrimental effects
on the deposition mechanism since the electrochemical features are
comparable.
The influence of the buffering agent on the variation of cathode
potential during deposition is more pronounced. Figure 12 displays a
set of E-t recordings obtained in nonbuffered acetonitrile-based
Ϫ2
mixtures for j varying between 0.25 and 4 mA cm . In the same
figure the extreme curves recorded in the buffered solution have
been superimposed. Figure 12 shows a large potential oscillation
whatever the current density in nonbuffered solution, which is vir-
tually absent or strongly attenuated when strontium hydroxide is
added. By contrast with deposition performed in pure DMSO solu-
tion, potential oscillations in the first hundred seconds of deposition
are not associated to the cogeneration of hydrogen bubbles of large
size. Indeed, bubbles are still present in the buffered solution despite
the absence of any oscillation; they are even causing the deposit to
peel off from the substrate. This is made clear in Fig. 12, where the
cathode potential is increasing for long deposition times because
growing parts of the deposits are stripped off from the silver elec-
trode. The potential oscillation might be attributed to an initial re-
pulsive interaction between the reactive species and silver, respon-
sible for a negative potential shift followed by an opposite positive
one after adsorption on the surface. Such an interpretation is consis-
tent with the electrochemical study, which showed the occurrence of
prepeaks at platinum and perturbations at silver electrodes on ca-
thodic polarization curves. This interaction is expected to disappear
when the metallic electrode is fully covered by the deposit, and
consequently the later evolution of the potential consists of a con-
tinuous decay as in pure dimethyl sulfoxide.
Figure 14. Surface micrographs of as-grown films deposited in 0.1 M
SrCl •6H O. Substrate: Ag. Solvent: acetonitrile ϩ DMSO. ͑a͒
j
2
2
Ϫ2
Ϫ2
ϭ 4 mA cm . Q ϭ 28 C. ͑b͒ j ϭ 0.25 mA cm . Q ϭ 28 C.
rounding rings of 1 ␮m diam ͑Fig. 14b͒. As mentioned before, films
deposited in buffered acetonitrile medium were not adherent and
therefore could not be examined further.
Microprobe analysis has shown contamination by chloride for all
films prepared in acetonitrile-based solvent; % At (Cl) ϭ 10 over
Ϫ2
Ϫ2
1 mA cm and % At (Cl) ϭ 5 below 1 mA cm . Thus the sur-
face aggregates should be associated with a higher chloride content
than in the rest of the deposit. X-ray imaging showed that the entire
surface contains chloride but did not allow demonstration of the
specific enrichment in chloride in later structures. Films were ana-
lyzed by XRD in order to check whether this contamination consists
of an intercalation of chloride ions within the predominant phase or
results from the presence of an additional phase. Figure 15 exhibits
Ϫ2
the XRD pattern obtained at 2 mA cm after formation and is
representative of the whole series in acetonitrile-based solution. Be-
sides the broad peaks related to SrCO and Sr(OH) which were
3
2
Potentials ranges are yet instructive to compare with one other;
in nonbuffered medium the cathode potential during deposition is
varying between Ϫ1.8 and Ϫ6 V, while in buffered medium the
lower limit has increased to Ϫ3 V, leaving the upper one un-
changed. The effective potential range is located in the reduction
wave for low currents and at the beginning of the plateau for higher
ones.
already present for films synthesized in pure DMSO electrolytic
solution ͑Fig. 10͒, the sharp set of additional peaks is undoubtedly
assignable to the monoclinic form of SrCl •2H O. Strontium chlo-
2
2
ride particles must be of larger size than those of strontium carbon-
ate, but their size is still nanometric. In fact both crystalline phases
are closely mixed since the chloride signal was detected everywhere
at the surface by the electron microprobe.
SEM observations indicate that rather different surface mor-
phologies are obtained in mixed solvent as compared to pure
Strontium chloride is weakly soluble in pure acetonitrile and also
scarcely soluble in the 6:1 mixture with dimethyl sulfoxide. Its con-
centration in the interfacial region should reach an oversaturation
value provided that the repartition of the solvent molecules is dif-
ferent from equilibrium in the solution volume. Concentration gra-
dients due to diffusion and convective effects are supposed to cause
spontaneous surface crystallization of the less soluble strontium
DMSO. Figure 14a shows a typical surface view of a film synthe-
Ϫ2
sized at high current density (4 mA cm
) in nonbuffered
acetonitrile-based solution. The layer is adherent to the silver sub-
strate but cracked. It is composed of a dense underlayer with short
needles pointing out from it and surmounted from place to place by
filamentous aggregates. When the deposition current is lowered, the
aggregates become smoother and turn into brushes of fibers sur-
compound, SrCl in its twofold hydrated variety. Chemical precipi-
2
tation of strontium chloride does not consume any electrical charge
Downloaded on 2015-06-14 to IP 134.129.120.3 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 150 ͑2͒ D30-D40 ͑2003͒
D39
SrCO was assessed by X-ray diffraction. We propose this reaction
3
to be the consequence of supersaturation of Sr͑II͒ chloride in the
vicinity of the cathode during deposition due to its low solubility in
acetonitrile. Therefore, the strontium precursor deposition involves
two nonequilibrium simultaneous pathways, on the one hand, the
usual electroprecipitation mechanism using electrogeneration of
base followed by precipitation of insoluble hydroxide; on the other
hand, crystallization of the Sr͑II͒ chloride salt under supersaturation
concentration.
CEA/Saclay assisted in meeting the publication costs of this article.
References
1
. J. Rouxel, M. Tournoux, and J. Brec, Soft Chemistry Routes to New Materials,
Trans Tech Publications, Switzerland ͑1993͒.
2
3
4
. G. H. Annal Therese and P. V. Kamath, Chem. Mater., 12, 1195 ͑2000͒.
. D. Tench and L. F. Warren, J. Electrochem. Soc., 130, 869 ͑1983͒.
. S. Gross, Review of Electrochemical Impregnation for Ni-Cd Cells, JPL, California
Institute of Technology, Pasadena, CA ͑1977͒.
5
6
. D. M. MacArthur, Electrochim. Acta, 15, 730 ͑1970͒.
. W. Fisher, Electrochim. Acta, 21, 1001 ͑1976͒.
Figure 15. XRD pattern of an as-grown film deposited in 0.1 M SrCl2
Ϫ2
7. D. A. Corrigan and S. L. Knight, Electrochim. Acta, 36, 613 ͑1989͒.
•
6H O. Solvent: acetonitrile ϩ DMSO. Substrate: Ag. j ϭ 2 mA cm .
2
8
. C. C. Streinz, A. P. Hartmann, S. Motupally, and J. W. Weidner, J. Electrochem.
Soc., 142, 1084 ͑1995͒.
Q ϭ 28 C.
9
. C. C. Streinz, S. Motupally, and J. W. Weidner, J. Electrochem. Soc., 142, 4051
͑
1995͒.
0. G. Fisher, Am. Ceram. Soc. Bull., 65, 283 ͑1986͒.
1. J. A. Switzer, Am. Ceram. Soc. Bull., 66, 1521 ͑1987͒.
2. Y. Zhou, R. J. Phillips, and J. A. Switzer, J. Am. Ceram. Soc., 78, 981 ͑1995͒.
13. I. Zhitomirsky and A. Petric, Mater. Lett., 40, 263 ͑1999͒.
1
1
1
while contributing to the final mass. This is a plausible explanation
for the excess mass observed with respect to 12.85 mg as expected
in the case of a pure SrCO phase, especially at low currents ͑Table
3
1
4. I. Zhitomirsky, L. Gal-Or, A. Kohn, and H. W. Hennicke, J. Mater. Sci., 30, 5307
1995͒.
III͒. The contribution of SrCl to the total mass is somewhat variable
2
͑
as a function of the deposition current; anyway relative diffraction
intensities of the strontium chloride and carbonate peaks indicate a
higher level under low current conditions, in good agreement with
mass measurements.
1
1
5. I. Zhitomirsky and L. Gal-Or, J. Eur. Ceram. Soc., 16, 819 ͑1996͒.
6. I. Zhitomirsky, Mater. Lett., 33, 305 ͑1998͒.
17. I. Zhitomirsky and L. Gal-Or, Mater. Lett., 31, 155 ͑1997͒.
18. L. Gal-Or, I. Silberman, and R. Chaim, J. Electrochem. Soc., 138, 1939 ͑1991͒.
1
9. R. Chaim, G. Shark, L. Gal-Or, and H. Bestgen, J. Mater. Sci., 29, 6241 ͑1994͒.
20. J. A. Switzer, M. J. Shane, and R. J. Phillips, Science, 247, 443 ͑1990͒.
1. R. G. Vadimsky, R. P. Frankenthal, and D. E. Thompson, J. Electrochem. Soc., 126,
017 ͑1979͒.
22. C. W. Nieh, E. Kolawa, F. C. T. So, and M.-A. Nicolet, Mater. Lett., 6, 177 ͑1988͒.
Conclusions
2
In this paper we have investigated the electroprecipitation
mechanism of strontium hydroxide from nitrate and chloride salts
dissolved in pure dimethyl sulfoxide and acetonitrile ϩ DMSO
mixtures in the presence of water as reactive species. The deposition
mechanism is understood as an electrochemically induced surface
precipitation of insoluble compounds mainly composed of metal hy-
droxide resulting from a local pH increase caused by the cathodic
formation of base. High-temperature sintering of the precursor film
leads to the final oxide product. We have confirmed that generation
of hydroxide ions proceeds mainly via water electrolysis. In a nitrate
medium, the contribution of the nitrate reduction reaction is minor,
although it is thermodynamically more favorable than water
reduction.
2
2
3. Q.-X. Jia and W. A. Anderson, IEEE Trans. Compon., Hybrids, Manuf. Technol.,
5, 121 ͑1992͒.
4. L. A. Bursill, I. M. Reaney, D. P. Vijay, and S. B. Desu, J. Appl. Phys., 75, 1521
1994͒.
25. M. Izaki and T. Omi, Appl. Phys. Lett., 68, 2439 ͑1996͒.
1
2
͑
2
2
2
6. S. Peulon and D. Lincot, J. Electrochem. Soc., 145, 864 ͑1998͒.
7. T. Pauport e´ and D. Lincot, Electrochim. Acta, 45, 3345 ͑2000͒.
8. D. Gal, G. Hodes, D. Lincot, and H.-W. Schock, Thin Solid Films, 361-362, 79
͑
2000͒.
29. H. Konno, M. Tokita, A. Furusaki, and R. Furuichi, Electrochim. Acta, 37, 2421
͑1992͒.
3
0. Y. Matsumoto, T. Morikawa, H. Adachi, and J. Hombo, Mater. Res. Bull., 27, 1319
1992͒.
͑
3
3
1. Y. Matsumoto, H. Adachi, and J. Hombo, J. Am. Ceram. Soc., 76, 769 ͑1993͒.
2. I. Zhitomirsky, A. Kohn, and L. Gal-Or, Mater. Lett., 25, 223 ͑1995͒.
A series of films has been galvanostatically deposited on silver
tapes as a function of current density, electrolyte concentration, and
nature of the solvent. An accurate control of these parameters does
exert a crucial influence on microstructural properties and can be
optimized to give good quality precursors. The optimum cathode
potential ranges between Ϫ2 and Ϫ3.5 V vs. SCE and corresponds
to the water reduction wave on cathodic polarization curves.
In pure DMSO, covering and adherent layers are obtained at low
33. I. Zhitomirsky, L. Gal-Or, and S. Klein, J. Mater. Sci. Lett., 14, 60 ͑1995͒.
3
3
4. I. Zhitomirsky, Am. Ceram. Soc. Bull., 79, 57 ͑2000͒.
5. I. Zhitomirsky, L. Gal-Or, A. Kohn, and H. W. Hennicke, J. Mater. Sci. Lett., 14,
8
07 ͑1995͒.
3
6. T. D. Golden, M. G. Shumsky, Y. Zhou, R. A. Vander Werf, R. A. Van Leeuwen,
and J. A. Switzer, Chem. Mater., 8, 2499 ͑1996͒.
37. J. Livage, M. Chatry, M. Henry, and F. Taulelle, Proc. MRS Symp., 271, 201
1992͒.
͑
3
3
8. L. Indira and P. V. Kamath, J. Mater. Chem., 4, 1487 ͑1994͒.
9. G. H. Annal Therese and P. V. Kamath, J. Appl. Electrochem., 28, 539 ͑1998͒.
Ϫ2
deposition currents, Ͻ2 mA cm in 0.1 M Sr͑II͒, and at higher
Ϫ2
values comprised between 5 and 15 mA cm in 0.5 M Sr͑II͒. The
40. R. N. Bhattacharya and R. D. Blaugher, Physica C, 229, 244 ͑1994͒.
4
1. H. Minoura, K. Naruto, H. Takano, E. Haseo, T. Sugiura, Y. Ueno, and T. Endo,
Chem. Lett., 3, 379 ͑1991͒.
films are well crystallized after formation. The major phase obtained
in chloride solution is not strontium hydroxide as in nitrate solution
but strontium carbonate, probably because of spontaneous reaction
of Sr(OH)2 with carbon dioxide. Electrodeposition from the chlo-
ride medium was demonstrated to be achieved with a faradaic effi-
ciency close to 100% but was found to be lower ͑70-80%͒ in the
nitrate solution. This point is also contradictory with the hypothesis
of an hydroxide-generating nitrate reduction reaction.
4
2. P. M. S. Monk, R. Janes, and R. D. Partridge, J. Chem. Soc., Faraday Trans., 93,
3
991 ͑1997͒.
43. R. N. Bhattacharya, R. D. Blaugher, Z. F. Ren, W. Li, J. H. Wang, M. Paranthaman,
D. T. Verebelyi, and D. K. Christen, Electrochem. Solid-State Lett., 1, 165 ͑1998͒.
4
4. K. A. Richardson, D. M. W. Arrigan, P. A. J. de Groot, P. C. Lanchester, and P. N.
Bartlett, Electrochim. Acta, 41, 1629 ͑1996͒.
4
5. V. N. Shinde and S. H. Pawar, Indian J. Phys., A, 68, 423 ͑1994͒.
46. F. L. Yang, Y. S. Chang, C. S. Li, S. M. Ma, Y. T. Huang, and W. H. Lee, J. Mater.
Sci., 27, 5739 ͑1992͒.
7. M. S. Mart ´ı n-Gonz a´ lez, J. Garc ´ı a-Jaca, E. Mor a´ n, and M. A. Alario-Franco, J.
Mater. Chem., 9, 137 ͑1999͒.
The acetonitrile ϩ DMSO solution has shown to be unsuited for
deposition since films exhibit poor adherence on silver substrates. In
this solvent, a second crystalline phase has been identified as stron-
´
4
4
8. P. Slezak and A. Wieckowski, J. Electrochem. Soc., 138, 1038 ͑1991͒.
tium chloride and coprecipitation of nanometric SrCl particles with
49. J. N. Butler, J. Electroanal. Chem., 14, 89 ͑1967͒.
2
Downloaded on 2015-06-14 to IP 134.129.120.3 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D40
Journal of The Electrochemical Society, 150 ͑2͒ D30-D40 ͑2003͒
5
5
0. P. Gendre, P. R e´ gnier, L. Schmirgeld-Mignot, and A. Marquet, Fr. Pat. 94 02710
54. B. Tr e´ millon, Electrochimie analytique et r e´ actions en solution, Vol. 2, p. 37,
Masson, Paris ͑1993͒.
55. Nonaqueous Electrochemistry, D. Aurbach, Editor, Marcel Dekker, Inc., New York
͑1994͒.
1. F. Legendre, P. R e´ gnier, L. Schmirgeld-Mignot, and S. Poissonnet, Eur. Phys. J.:
Appl. Phys., 1, 191 ͑1998͒.
͑1999͒.
5
5
2. G. Sch o¨ ber and V. Gutmann, Monatsch., 90, 897 ͑1959͒.
56. F. Portemer, A. Delahaye-Vidal, and M. Figlartz, J. Electrochem. Soc., 139, 671
͑1992͒.
3. E. L. Johnson, K. H. Pool, and R. E. Hamm, Anal. Chem., 38, 183 ͑1966͒.
Downloaded on 2015-06-14 to IP 134.129.120.3 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).