Electrodeposition and characterization of manganese coatings

Article abstract of DOI:10.1149/1.1452117

Manganese coatings of high quality are electrodeposited on steel substrates from simple sulfate solutions with addition of ammonium sulfate. Potentiodynamic scans and galvanostatic experiments are used to study manganese electrodeposition in a wide range of pH and current density. The effect of these variables on the microstructure, crystallography, mechanical, and corrosion-resistance properties of manganese deposits are investigated. It is found that ammonium sulfate enhances the reduction reaction of the manganese ion and provides a buffering effect. Two types of manganese deposits can be obtained depending on current density: crystalline films (type I, body-centered tetragonal γ-Mn) at low current density and amorphous films (type II) at high current density. Bright manganese films with (002) preferential orientation are electrodeposited at low pH. Type I structures show recrystallization at room temperature with phase transformation; the rate of phase transformation from γ-Mn to α-Mn (body-centered cubic) follows a Johnson-Mehl-Avrami kinetics. Crystalline films obtained at relatively high current density and low pH tend to have higher phase transformation rates. Amorphous films show good corrosion resistance both in acidic sodium sulfate/borate and sodium chloride electrolytes.

Full text of DOI:10.1149/1.1452117

Journal of The Electrochemical Society, 149 ͑4͒ C209-C217 ͑2002͒
C209
0013-4651/2002/149͑4͒/C209/9/$7.00 © The Electrochemical Society, Inc.
Electrodeposition and Characterization of Manganese Coatings
,z
Jie Gonga, and Giovanni Zangari
*
a,b,**
b
aMaterials Science Program and Department of Metallurgical and Materials Engineering,
University of Alabama, Tuscaloosa, Alabama 35487, USA
Manganese coatings of high quality are electrodeposited on steel substrates from simple sulfate solutions with addition of
ammonium sulfate. Potentiodynamic scans and galvanostatic experiments are used to study manganese electrodeposition in a wide
range of pH and current density. The effect of these variables on the microstructure, crystallography, mechanical, and corrosion-
resistance properties of manganese deposits are investigated. It is found that ammonium sulfate enhances the reduction reaction of
the manganese ion and provides a buffering effect. Two types of manganese deposits can be obtained depending on current
density: crystalline films ͑type I, body-centered tetragonal ␥-Mn͒ at low current density and amorphous films ͑type II͒ at high
current density. Bright manganese films with ͑002͒ preferential orientation are electrodeposited at low pH. Type I structures show
recrystallization at room temperature with phase transformation; the rate of phase transformation from ␥-Mn to ␣-Mn ͑body-
centered cubic͒ follows a Johnson-Mehl-Avrami kinetics. Crystalline films obtained at relatively high current density and low pH
tend to have higher phase transformation rates. Amorphous films show good corrosion resistance both in acidic sodium sulfate/
borate and sodium chloride electrolytes.
© 2002 The Electrochemical Society. ͓DOI: 10.1149/1.1452117͔ All rights reserved.
Manuscript submitted March 15, 2001; revised manuscript received October 26, 2001. Available electronically March 4, 2002.
Electrodeposited coatings of manganese ͑Mn͒ or Mn alloys po-
tentially combine good sacrificial corrosion protection with adequate
tribological behavior and suitable mechanical properties for coating
steel products.1-4 Despite this interest, few investigations have been
concerned with a detailed study of the electrodeposition of pure Mn
metal and the structure and properties of its deposits. This is prob-
ably due to the very low potential of Mn in aqueous solutions
(E0(Mn2ϩ/Mn) ϭ Ϫ1.18 VSHE), which renders Mn the most elec-
tronegative metal that can be electrodeposited from aqueous solu-
tions, leaving a narrow processing window available for optimiza-
tion of the deposition process.
Although manganese electrowinning has been an industrial pro-
cess since 1930, recent literature on the electrodeposition of manga-
nese and the relationship between its physical properties and pro-
cessing conditions is rather sparse. Dean’s book5 provides an
excellent survey of the status of manganese electrowinning technol-
ogy up to 1952, and Brenner’s book1 offers an exhaustive review of
manganese and Mn alloy electroplating up to 1960. Early literature
focused on electrowinning and consequently studied mainly the ef-
fect of bath composition, pH, current density, etc., on current effi-
ciency, coating appearance, and power consumption.6-10 Manganese
could be electrodeposited from both sulfate and chloride solutions,
with or without additives. A current efficiency of up to 60-70% and
both ductile gamma manganese ͑body-centered tetragonal, bct, or
more precisely, centered tetragonal11͒ and brittle alpha manganese
͑body-centered cubic, bcc͒ could be obtained under different condi-
tions. The ductile metal was found to recrystallize to the brittle form
at room temperature, the transformation being complete in about 14
days.12 This phenomenon was ascribed to the high hydrogen content
of electrolytic Mn.13 The phase transition kinetics was investigated
through the evolution of electrical resistivity with time.12 By using
production of high-quality Mn deposits, which could be taken as a
starting point in the development of viable processes for the elec-
trodeposition of various Mn alloys.
Experimental
The electrodeposition experiments were carried out in a three-
electrode, two-compartment prismatic cell. The counter electrode
was a platinum foil placed in the same compartment as the working
electrode, and covered by a diaphragm to avoid the contamination of
solution following formation of manganese oxides. The saturated
calomel reference electrode ͑SCE͒ was in a different compartment,
separated from the working electrode by a Luggin capillary mounted
on a syringe barrel. The same cell was used for potentiodynamic
scans and galvanostatic electrodeposition. The experiments were
performed, and potential/current curves as well as galvanostatic
transients were recorded, using an EG&G PAR A273 potentiostat/
galvanostat.
Simple sulfate electrolytes with addition of ammonium sulfate
were used. pH was adjusted by adding concentrated ammonium hy-
droxide or sulfuric acid. No attempt was made to control the pH
during deposition, but pH was measured and, if necessary, adjusted,
after each deposition experiment. The basic bath contained 0.59 M
MnSO4 and 1 M (NH4)2SO4 . The solutions were prepared with
analytical grade reagents and triply distilled water. All experiments
were carried out at 25°C.
Manganese was plated on stainless steel 304 plates with an active
surface area of 3 cm2 (1.5 ϫ 2.0 cm). The substrates were first
mechanically polished with various grades of alumina powder pa-
pers and degreased with alkaline solution ͑23 g/L NaOH, 22 g/L
Na2CO3 , 10 g/L Na2SiO3(anhydrous), 10 g/L Na3PO4 , 1 g/L so-
dium lauryl sulfate, 90-95°C, 3-5 min͒ and acetone. Subsequently,
they were electropolished in concentrated phosphoric acid ͑85%͒
and pickled in mixed nitric ͑5%͒ and hydrochloric acid ͑25%͒ just
before use.
modern electrochemical techniques, more recent papers14-18 exam-
ined the influence of processing conditions on manganese elec-
trodeposition.
The current efficiency was calculated from the weight gain of the
sample. Surface morphology was examined by scanning electron
microscopy ͑SEM͒ using a Philips XL30 instrument. Chemical
analysis of the deposits was performed by energy-dispersive X-ray
spectrometry ͑EDAX͒ attached to the SEM, while the incorporation
of oxygen and the chemical states of Mn were studied by X-ray
photoelectron spectroscopy ͑XPS͒ using a Kratos Axis 165. Crystal-
lographic structure was determined by X-ray diffraction ͑XRD͒ us-
ing a Philips APD 3520 diffractometer with Cu K␣ radiation. A
Buehler Micromet microhardness tester ͑model 1600-6100͒ was em-
ployed to measure the microhardness of manganese films.
Corrosion resistance of the coatings was evaluated by monitoring
the open-circuit potential of the deposits after immersion in
0.5 M Na2SO4 ϩ 0.5 M H3BO3 ͑pH 3͒ or 2.5% NaCl ͑pH 3͒ solu-
In this paper, the electrodeposition of Mn from ammonium sul-
fate electrolytes at pH 1.7-7.5 and current density from 30 to 330
mA/cm2 is studied. Morphological, chemical, structural character-
ization, and open-circuit potential measurements were further used
to characterize the properties of the deposits. In order to attempt
stabilization of ductile gamma manganese, the phase transformation
kinetics and its dependence on current density and pH were inves-
tigated. This work provides a range of processing conditions for the
* Electrochemical Society Student Member.
** Electrochemical Society Active Member.
z E-mail: gong002@bama.ua.edu
Downloaded on 2014-10-06 to IP 131.215.225.9 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C210
Journal of The Electrochemical Society, 149 ͑4͒ C209-C217 ͑2002͒
Figure 2. Effect of current density on current efficiency at pH 6.4, 0.59 M
MnSO4 and 1 M (NH4)2SO4 .
with addition of ammonium sulfate may indicate a different mecha-
nism for Mn and/or hydrogen reduction. The scan rate has little
influence on the potentiodynamic behavior until a relatively high
overvoltage is attained, where a low scan rate yields a higher current
density at the same overpotential. This is contrary to the common
trend of voltammetric measurements and can be explained as fol-
lows: as the potentiodynamic scan proceeds, the stainless steel cath-
ode would be completely covered by manganese at lower potentials
when the scan rate is lower. Manganese has a higher catalytic activ-
ity for Mn2ϩ discharging and a slightly higher overpotential for Hϩ
discharging than stainless steel,17,19 causing observation of a larger
current. This effect was confirmed by galvanostatic depositions.
At all pH values investigated, Mn͑II͒ discharge is accompanied
by hydrogen evolution. At pH 2-3 a lower overvoltage for hydrogen
evolution is observed ͑Fig. 1b͒. The consequent increase of hydro-
gen evolution and the fast dissolution of the Mn coating in acid
solutions limit the practical pH range for deposition above these
values. A manganese electrodeposition peak can be seen around
Ϫ1.7 VSCE , corresponding to a current density of about 25-30
mA/cm2 ͑see inset in Fig. 1b; note the log scale of the vertical axis͒.
To summarize, the Mn͑II͒ ion discharging reaction is accompanied
by hydrogen evolution at all pH values, while ammonium sulfate
increases manganese ion discharging ability and provides a desirable
buffering effect at pH 2-3.5 and 6-7.
Figure 1. Potentiodynamic behavior of Mn 0.59 M at room temperature: ͑a͒
pH 3, with and without 1 M (NH4)2SO4 ; ͑b͒ with (NH4)2SO4 , at various
pH values. Note the vertical log axis of the inset.
tions. The solutions were quiescent and no inert gas was used to
purge the electrolyte. Morphology and chemical composition of the
coatings after corrosion testing were evaluated to gain insight into
the extent and nature of the corrosion process.
Results and Discussion
Potentiodynamic behavior.—The pH of the as-prepared electro-
plating solution was 3.7, which was adjusted in the range from 1.7
to 7.5 by addition of concentrated ammonium hydroxide or sulfuric
acid. As pointed out by Dean,5 the addition of ammonium sulfate to
the manganese electrolyte can prevent precipitation of manganese
hydroxides during electrodeposition as the pH increases and im-
proves the conductivity of the solution. In addition, we observed a
buffering effect of the ammonium sulfate at pH above 6 and below
3.5. In the presence of air, at high pH ͑7-9͒, the electrolyte was not
stable unless sulfur compounds or other antioxidants were added,
such as sulfur dioxide and hydroxylamine salts.5 These substances,
however, are generally undesirable from the viewpoint of corrosion
protection; consequently, simple sulfate acidic electrolytes were pre-
ferred in this work.
Figure 1 reports the cathodic current/voltage curves for the Mn
electrolyte, with and without addition of ammonium sulfate ͑Fig.
1a͒, and at various pH ͑Fig. 1b͒. The increase in current density
below Ϫ1.5 VSCE with addition of ammonium sulfate observed in
Fig. 1a is due both to an increase in the discharging ability of Mn
ions, as reported in Ref. 5, and to a decrease in the overpotential for
hydrogen evolution.17 As demonstrated by galvanostatic deposition,
the presence of ammonium sulfate is essential to grow manganese
coatings with good coverage. The increase in the slope of the curve
Galvanostatic electrodeposition.—The manganese deposits were
grown galvanostatically at a temperature of 25°C, with current den-
sity varying in the range 30-330 mA/cm2 and pH from 1.7 to 7.5.
Deposition was carried out without stirring of the electrolyte, for a
time duration calculated to obtain a 10 ␮m thick deposit according
to Faraday’s law, assuming 100% current efficiency and the deposits
in the tetragonal ␥-form (␳ ϭ 7.20 g/cm3). The actual current effi-
ciency was calculated from the weight gain of the cathode. The
results are reported in Fig. 2 as a function of current density for
fixed pH 6.4. The highest current efficiency is about 60% at the
current density of 65 mA/cm2. This is very close to the results re-
ported by Dean5 at pH 8.2-8.4. At low current densities, the effi-
ciency of manganese deposition is affected by self-dissolution of the
film. Because both hydrogen ion discharging
2Hϩ ϩ 2e → H2↑
͓1͔
and manganese dissolution
Mn ϩ 2Hϩ → Mnϩ ϩ H2↑
͓2͔
contribute to hydrogen evolution, the hydrogen evolution rate is
high and the manganese deposition rate appears low. A sharp fall of
Downloaded on 2014-10-06 to IP 131.215.225.9 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, 149 ͑4͒ C209-C217 ͑2002͒
C211
Figure 3. Effect of pH on current efficiency ͑current density 65 mA/cm2͒,
0.59 M MnSO4 and 1 M (NH4)2SO4 .
Figure 4. High-resolution XPS spectra of manganese coatings in the O 1s
and Mn 2p energy windows: ͑top͒ type I, 65 mA/cm2 and ͑bottom͒ type II,
330 mA/cm2.
current efficiency when current density was increased from 65 to
150 mA/cm2 was also observed, in agreement with both earlier5 and
recent results16,18 in sulfate and chloride solutions. Various reports
indicate different values of this transition current density, which
might be due to the different deposition conditions employed.
For aqueous electrolytes containing MnSO4 and (NH4)2SO4 ,
MnO͑OH͒ also makes the solution unclear and lowers current effi-
2
ciency, which is detrimental to manganese deposition.
2ϩ
2ϩ
Mn2ϩ can exist in the forms Mn(H O)
,
Mn(NH )
,
͓
͔
͓
͔
xϭ1-6
2
6
3
20-22
2ϩ
Microstructure and composition of the coatings.—Adherent
manganese deposits of good quality are generally obtained by this
process. At current densities from 30 to 100 mA/cm2, coatings are
silvery matte in appearance, compact, uniform, and continuous;
these deposits are indicated in the following discussion as type I.
Typical microhardness of as-deposited type I coatings is 147 Knoop
hardness number ͑KHN͒, using a load of 200 gf on a 40 ␮m thick
film obtained at the current density of 65 mA/cm2. The depth of
indent is about 11% of the deposit thickness. On the other hand,
black, glossy, and even more compact films ͑which are called type II
in the following discussion͒ are formed at current densities above
150 mA/cm2. Microhardness of type II deposits is typically 487
KHN, using a load of 200 gf on 40 ␮m thick films obtained at the
current density of 330 mA/cm2. The depth of indent is about 6.3% of
the deposit thickness. At current densities between 100 and 150
mA/cm2 a transition between type I and type II films is observed;
occasionally the two microstructures are present in the same sample,
with type I observed at the center area and type II at the edges.
The chemical states and composition of typical type I and type II
coatings were examined by EDAX and XPS depth profiling. Al-
though EDAX seems to indicate that both types of films are pure Mn
and do not contain other impurities, XPS analysis gives more accu-
rate results. Figure 4 shows high-resolution XPS scans of the Mn 2p
and O 1s regions for the two types of films at different depths. For
type I films, the Mn:O ͑atom %͒ ratio changes from 51:49 near the
surface ͑after etching five cycles, about 10 nm deep͒ to 85:15 at the
interior ͑after etching 40 cycles, about 80 nm deep͒. By deconvolut-
ing the O 1s peak using three components ͑corresponding to water,
oxide, and hydroxide͒,23 the existing forms of oxygen could be
quantified. Analyzing the Mn 2p peaks with the charge neutralizer
on to avoid sample charging effects, Mn oxidation states could be
determined. Near the surface, manganese is almost exclusively in
the form Mn2ϩ to Mn4ϩ, probably as MnOx(OH)2Ϫx (x ϭ 0-2),
while in the bulk of the deposit, Mn is mostly present as Mn0. For
type II coatings, composition and chemical states do not depend on
depth. Mn is mostly in the ϩ2 and ϩ3 valence states. Elemental
manganese content is as low as 6-8 atom %. XPS results thus con-
firmed the depositing mechanism for the proposed type II coatings.
SEM surface micrographs of samples prepared at current density
30-330 mA/cm2, pH 6.4, are shown in Fig. 5. Regularly shaped
or Mn(H O)6Ϫn(NH3)n
.
Due to the presence of the high-
͓
͔
2
spin 3d5 manganese͑II͒ ion, all these compounds are octahedrally
coordinated and possess zero ligand field stabilization energy.20
Therefore, the complexes formed in this system are unstable. It was
also proposed22 that Mn2ϩ discharge could take place through the
following mechanism when NH3 is present
Mn H O͒
NH ͒ 2ϩ ϩ xH2O
͔
3 n
͓
͑
͑
2
6Ϫn
2ϩ
↔
Mn H O͒
ϩ nNH3
͓3͔
͓
͑
͔
6Ϫnϩx
2
Mn H O͒
2ϩ ϩ 2e → Mn ϩ 6 Ϫ n ϩ x͒H O ͓4͔
͓
͑
͔
͑
2
6Ϫnϩx
2
Due to the relative instability of the possible Mn2ϩ complexes, as
current density increases from 65 to 150 mA/cm2, it is unlikely that
the dissociation of these complexes becomes the rate-determining
step of Mn2ϩ discharge. It is more probable that Mn2ϩ discharging
is limited by the diffusion of Mn2ϩ complexes to the cathode
surface.16 When such limiting current is attained, any further in-
crease of current density would only increase the Hϩ discharging
rate, decreasing the current efficiency and eliminating the Hϩ adja-
cent to the cathode surface. The buffering effect of (NH4)2SO4
could stabilize the pH near the cathode only to a limited extent. With
increasing current density the hydroxyl ion concentration at the in-
terface would increase to such a high level that large amounts of
manganese hydroxide would precipitate and be incorporated into the
growing deposit, increasing the apparent current efficiency as cur-
rent density goes above 150 mA/cm2 ͑Fig. 2͒. Incorporation of Mn
hydroxides at high current density was confirmed by XPS analysis,
as discussed in the next section.
Current efficiency vs. pH at a fixed current density of 65 mA/cm2
is plotted in Fig. 3. Current efficiency is relatively stable at above
70% from pH 2 to 6. Current efficiency decreases below pH 2, at
which the high electrolyte acidity and the increased Hϩ discharging
rate induce higher Mn dissolution rate and lower relative Mn depo-
sition rate, respectively. From pH 2 to 4, Mn2ϩ exists predominantly
as Mn(H O) 2ϩ, and at pH 4 to 6 as Mn(H O)6Ϫn(NH3)n 2ϩ, n
͓
͔
͓
͔
2
6
2
increasing with pH.20-22 Above pH 6.5, in the presence of air, the
bulk electrolyte becomes gradually unstable due to the precipitation
of manganese hydroxide Mn͑OH͒
. Its oxidation product
͓
͔
2
Downloaded on 2014-10-06 to IP 131.215.225.9 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C212
Journal of The Electrochemical Society, 149 ͑4͒ C209-C217 ͑2002͒
Figure 5. SEM surface micrographs of manganese electrodeposited at various current densities and pH 6.4.
crystalline grains with apparent grain size in the range 3-6 ␮m are
formed at the current density from 30 to 100 mA/cm2 ͑type I depos-
its͒. Good coverage can be obtained already at low current density
͑30-40 mA/cm2͒; these films are composed of small ͑about 3 ␮m͒
and uneven grains ͑Fig. 5, 30 mA/cm2͒. The relatively low grain size
may be induced by the comparatively large hydrogen evolution rate.
The grain size increases to about 6 ␮m at 60 mA/cm2, where current
efficiency is maximum, then decreases gradually as current density
goes up to 100 mA/cm2 ͑Fig. 5, 30-100 mA/cm2͒. The latter trend is
reasonable on account of the current density dependence on nucle-
ation rate. Between 100 and 150 mA/cm2 a transition between the
type I and II deposits is observed. The corresponding films are not
continuous ͑Fig. 5, 120 mA/cm2͒, indicating difficult nucleation or
poor adhesion of the type I structure in the presence of type II,
which incorporates a large amount of Mn hydroxide precipitates. As
current density increases above 150 mA/cm2, initially a globular
structure and then a very fine-grained fibrous microstructure
(50-100 nm ϫ 100-500 nm), randomly oriented ͑Fig. 5, 200
mA/cm2͒ or aligned perpendicularly to the substrate ͑Fig. 5, 250
mA/cm2͒ are observed. The black and glossy appearance is probably
due to the fibrous microstructure.
Manganese films were also electrodeposited at different pHs,
from 1.7 to 7.5, at 65 mA/cm2 current density ͑where type I deposits
are obtained͒. SEM surface micrographs of the resulting deposits are
shown in Fig. 6. Sound and homogeneous deposits can be obtained
from pH 2 to 7. Below pH 2, discontinuous films are obtained as a
consequence of strong hydrogen evolution and Mn dissolution. As
pH decreases from 7, appearance of the deposits changes gradually
from silvery matte to silvery bright, the transition taking place from
pH 4 to pH 3. This corresponds to a transition of the microstructure
from clearly defined small grains to large grains, seemingly faceted
or composed of several subgrains ͑Fig. 6, pH 6.1-3.0͒. As the con-
2ϩ
centration of Mn(NH )
complexes increases with pH, a
͓
͔
3
x
smaller grain size at higher pH could be due to the increasing over-
voltage for metal reduction, as shown in Fig. 1b.21,24 The grains
formed at low pH are separated by sharply contrasted boundaries,
while inside each grain, a lamellar structure oriented along a specific
direction is observed ͑Fig. 6, pH 2.1͒. This microstructure is a pos-
Downloaded on 2014-10-06 to IP 131.215.225.9 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, 149 ͑4͒ C209-C217 ͑2002͒
C213
Figure 6. SEM surface micrographs of manganese electrodeposited at different pH values, current density 65 mA/cm2.
sible reason for the brightness of these deposits.24 The latter micro-
structure is textured, as confirmed by XRD characterization in the
following section.
amounts of impurities codeposited with the metal, which are known
to greatly inhibit crystal growth.24 No ␣-Mn could be observed for
the freshly deposited coatings at all current densities observed.
Small diffraction peaks due to a ␥-Mn2O3 phase are observed. The
formation of ␥-Mn2O3 could be attributed to the high chemical
activity of freshly deposited manganese.
Crystallographic structure and phase transformation.—XRD
was performed to determine the crystalline structure of the deposits.
XRD patterns ͑␪-2␪ scans͒ of the deposits obtained at different cur-
rent densities are shown in Fig. 7. At pH 6.4, pure ␥-Mn is obtained
at current density from 30 to 100 mA/cm2 ͑type I͒. With increasing
current density, the main diffraction peaks become broader, indicat-
ing a decrease in grain size or a transition to an amorphous structure
͑type II͒. This observation is in agreement with microstructural evi-
dence from SEM micrographs and can be explained in terms of large
Room-temperature recrystallization with phase transformation is
observed for the type I microstructure, with the ductile ␥-Mn gradu-
ally transforming to brittle and hard ␣-Mn as shown in Fig. 8. This
confirms the results reviewed in Ref. 1, although in our case no
purification of the electroplating bath was necessary to obtain the
␥-phase. Microhardness of type I deposits increased during the
transformation from 147 KHN to 583 KHN as the atomic fraction of
␣-Mn increased ͑Fig. 9͒. For type II coatings, both XRD ͑Fig. 10͒
and microhardness ͑Fig. 9͒ do not show any significant changes. On
account of the overlapping of the main peaks of ␣-Mn ͑411͒ and
␥-Mn ͑101͒, a quantitative analysis to determine the proportion of
the ␣ and ␥ phases and the preferential orientation of the films
would be unreliable.16 We roughly estimated the relative fraction of
the ␣ and ␥ phases by two methods: ͑i͒ calculation of the relative
intensity of the ␣ ͑411͒ and ␥ ͑101͒ peaks, and ͑ii͒ monitoring the
variation of microhardness with aging time. With regard to the latter
method, by assuming that the hardness of the coating is given by a
weighed average of the hardness of the two phases, hardness should
be a reliable estimate of the volume ratio of the two phases. By
using both methods it was found that the phase transformation ki-
netics closely followed the standard Johnson-Mehl-Avrami equation
f ϭ 1 Ϫ exp Ϫktn͒
͑
where f is the volume fraction transformed, t is time, k is a constant,
and n is the Avrami exponent.25 The results are shown in linearized
form log10 ln(1/(1 Ϫ f ) vs. log10 t in Fig. 11, together with data of
electrical resistance vs. time taken from Ref. 12. All the linear fits
Figure 7. XRD patterns of as-deposited Mn films at pH 6.4. Note the broad-
ening of the ␥-Mn diffraction peaks with increasing current density.
Downloaded on 2014-10-06 to IP 131.215.225.9 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C214
Journal of The Electrochemical Society, 149 ͑4͒ C209-C217 ͑2002͒
Figure 10. XRD patterns at different aging times for type II films electrode-
posited at pH 6.4 and current density 300 mA/cm2. Note the constant width
of the main diffraction peak.
exhibit higher transformation rates than those obtained at low cur-
rent density ͑Fig. 12͒. This cannot be explained on the basis of
microstructural data, assuming that grain boundaries are preferential
sites for nucleation of the new phase. In fact, the grain size and grain
boundary density of the samples grown at 30 mA/cm2 are compa-
rable with those of 100 mA/cm2 samples, and comparable transfor-
mation kinetics should be expected in the two cases. This implies
that defects other than grain boundaries are responsible for the trans-
formation. Gamali et al.26 proposed an explanation for the observed
dependence of transformation kinetics on current density in biphasic
␣ ϩ ␥ deposits obtained at high current densities. In the region of
the limiting current density for manganese deposition, the ␥-Mn
lattice structure is highly distorted due to hydrogen incorporation,
and nuclei of the ␣-Mn modification are formed preferentially in the
stressed, high-energy regions. This line of reasoning seems to be
applicable also to our case, though no ␣-Mn can be detected by
XRD of freshly deposited samples ͑Fig. 7͒. According to Ref. 24,
the reduced hydrogen diffusing into the deposit while plating is
responsible for the internal stresses by causing a temporary lattice
expansion in the deposited metal. During deposition hydrogen can
diffuse into the base metal as well as back into the electrolyte. At the
Figure 8. XRD patterns taken at various time intervals after deposition,
showing the phase transformation of ␥- to ␣-Mn. Electrodeposition at pH 6.4
and current density 65 mA/cm2.
gave correlation coefficients above 96% and yielded Avrami expo-
nents of 3.3-3.5. For a polymorphic transformation, that is with no
change of composition, and for a nucleation and growth process, the
theory predicts an Avrami exponent of 3 for an instantaneous, 3D
nucleation and growth process, or an exponent of 4 for a progres-
sive, 3D nucleation and growth process. An exponent between 3 and
4 is more realistic, as both instantaneous and strictly progressive
nucleation are idealized models. Such values of the Avrami expo-
nent can be interpreted in terms of a transformation starting at defi-
nite defect sites, with a nucleation rate decreasing with time as the
number of available sites decreases.
The transformation rates are dependent on the growth conditions
of the Mn deposits. Type I deposits obtained at high current density
Figure 11. Linearized plot of the relationship between the volume fraction
of ␥ phase transformed ͑W͒ and aging time ͑t͒. Linear fits give K and n
values from microhardness, XRD, and resistivity12 data. From microhard-
ness: K ϭ 10Ϫ19.9, n ϭ 3.42; from XRD data: K ϭ 10Ϫ19.2, n ϭ 3.36;
from resistivity data: K ϭ 10Ϫ20.79, n ϭ 3.47.
Figure 9. Microhardness vs. aging time at room temperature for type I and
II Mn deposits.
Downloaded on 2014-10-06 to IP 131.215.225.9 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, 149 ͑4͒ C209-C217 ͑2002͒
C215
Figure 12. XRD patterns of type I films obtained at various current densi-
ties, after 7 days. The rate of phase transformation increases with current
density.
Figure 14. Relative integrated intensities of the diffraction peaks for as-
deposited ␥-Mn as a function of pH.
Ii/Ii0
Iir ϭ
end of the deposition process, however, any hydrogen present in the
lattice beyond the solubility limit diffuses out of the metal; conse-
quently, its lattice shrinks and tensile stresses arise. A large amount
of hydrogen is evolved during Mn electrodeposition, and it is rea-
sonable to assume that the hydrogen content in manganese deposits
is always at its saturation level13 and does not change with current
density. The lattice shrinkage and the strain energy of the Mn lattice
would thus be related to the amount of supersaturated hydrogen
temporarily incorporated during electrodeposition, which increases
with current density. The difference in stored strain energy may thus
account for the difference in the phase transformation kinetics. The
observed shift of the ␥-Mn diffraction peaks to higher angles ͑lower
lattice parameters͒ with increasing current density ͑Fig. 7͒ can be
related to this effect.
The crystalline structure of manganese electrodeposited at vari-
ous pH values at a fixed current density of 65 mA/cm2 was also
examined by XRD ͑␪-2␪ scan͒. The results are shown in Fig. 13.
The deposits obtained at low pH ͑less than 3.7͒ exhibit a strong
͑002͒ preferential orientation, as do those obtained at pH 7.0. Figure
14 shows the relative intensity, Ir , of the diffraction peaks of ␥-Mn
as a function of pH. The relative intensity is defined as27
1
Ii/Ii ͒
͑
͚
0
n
where Ii and Ii0 represent experimental and standard28 integrated
peak intensities of each main peak, respectively, and n is the number
of the peaks considered. For a randomly oriented specimen, the
relative intensity of each peak should be equal to 1. For a certain
reflection i, a Iir value larger than 1 indicates that the corresponding
crystallographic direction is preferred, while a value lower than
1 shows that the corresponding orientation is ‘‘preferentially
avoided.’’ 29 Figure 14 shows that the extent of preferential orienta-
tion of ͑002͒ presents a minimum around pH 4.5 and increases at
high and low pH values.
Evaluation of the phase transformation kinetics of samples ob-
tained at different pH ͑Fig. 15͒ showed that Mn deposits obtained at
lower pH undergo a faster phase transformation than those obtained
at higher pH. This could also be explained by internal stresses in-
duced by hydrogen incorporation in the Mn film. Hydrogen dis-
charge rate at low pH increases, and consequently, higher internal
stresses would be developed and larger strain energy may be stored
in manganese coatings obtained at low pH.
Figure 13. XRD patterns of as-deposited Mn at the current density of 65
mA/cm2 and various pH values. Note the change of the ͑002͒ peak intensity
with decreasing pH.
Figure 15. XRD patterns of electrodeposited Mn grown at different pH after
4 days. The rate of phase transformation is higher at lower pH.
Downloaded on 2014-10-06 to IP 131.215.225.9 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C216
Journal of The Electrochemical Society, 149 ͑4͒ C209-C217 ͑2002͒
Figure 17. SEM micrographs of manganese coatings after corrosion testing
in 2.5% NaCl at pH 3.0: ͑a, left͒ type I, after 10,800 s immersion and ͑b,
right͒ type II, after 21,600 s immersion.
This oxide surface appears black and dull, and in some cases it was
already formed on the as-plated samples, as indicated by XPS analy-
sis. Such oxide-hydroxide film could act as a passive film capable of
maintaining the galvanic potential while acting as a kinetic barrier
for corrosion processes. This behavior may be similar to that of
some steels, when iron in the presence of certain alloying elements
can form Fe-hydroxides, which provide stable passive films.
For comparison, 1 ␮m thick cadmium coatings electroplated on
steel from commercial electrolytes were tested in the same solutions,
and results are shown in Fig. 18. In sodium sulfate/borate solutions
the transition time of cadmium is longer but comparable to that
observed for Mn coatings. Sacrificial protection in chloride solu-
tions, however, seems to last much longer for Mn coatings.
Figure 16. Ecorr vs. time plots for electroplated manganese immersed in ͑a͒
0.5 M Na2SO4 ϩ 0.5 M H3BO3 at pH 3.0 and ͑b͒ 2.5% NaCl at pH 3.0.
Conclusions
Manganese coatings of high quality can be electroplated from
sulfate solutions with addition of ammonium sulfate. Ammonium
sulfate improves manganese ion discharging ability and provides a
buffering effect.
Current density has an important influence on current efficiency,
appearance, microstructure, crystal structure, and corrosion resis-
tance of manganese deposits. ␥-Mn ͑bct, type I͒ can be formed at
current densities from 30 to 100 mA/cm2, while amorphous coatings
͑type II͒, containing large amounts of oxyhydroxides, can be
To summarize, a ␥ to ␣ phase transformation takes place upon
aging at room temperature in crystalline but not in amorphous de-
posits. The transformation can be slowed down by electrodepositing
at low current density and high pH.
Corrosion-resistance measurements.—Corrosion tests were per-
formed in sodium sulfate/borate and chloride electrolytes on both
type I and type II coatings. The test samples were about 6 ␮m thick.
Figure 16 shows plots of corrosion potential Ecorr vs. time for
type I and II manganese deposited at pH 6.4, 65 and 330 mA/cm2,
respectively. The corrosion performance of the two types of coating
is very different. In the sodium sulfate/borate system ͑Fig. 16a͒,
taking E ϭ Ϫ1.6 VSCE as the corrosion potential Ecorr of manganese
͑this plateau corresponds to manganese dissolution from the sub-
strate͒ and E ϭ Ϫ0.8 VSCE the voltage observed when dissolution is
completed, the performance of the amorphous coating is much bet-
ter. Ecorr of amorphous coatings in fact increases gradually toward
the value of the substrate, while the change in Ecorr for the crystal-
line coatings is more abrupt and takes place after a shorter time.
In the sodium chloride solution ͑Fig. 16b͒, corrosion resistance
of both coatings seems to be better, as Ecorr remains below Ϫ1 VSCE
in the time investigated. The surface of the coatings after the corro-
sion test was analyzed by SEM and EDAX ͑Fig. 17͒. The morphol-
ogy of amorphous coatings observed after a 21,600 s test appears to
be less deteriorated than that of crystalline coatings observed after a
test of duration 10,800 s. In the former case ͑Fig. 17b͒, no substrate
area is exposed as evidenced by EDAX. Furthermore, a layer with
an apparent composition of MnO2 is formed on the coating surface,
as confirmed by XPS data, which could slow corrosion kinetics.
Figure 18. Ecorr vs. time plots for electroplated cadmium ͑thickness 1 ␮m͒
in 0.5 M Na2SO4 ϩ 0.5 M H3BO3 at pH 3.0 and 2.5% NaCl at pH 3.0.
Downloaded on 2014-10-06 to IP 131.215.225.9 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, 149 ͑4͒ C209-C217 ͑2002͒
C217
formed at higher current densities, above 150 mA/cm2. Room-
temperature recrystallization of type I samples is observed, with
phase transformation from ␥-Mn ͑bct͒ to ␣-Mn ͑bcc͒; the kinetics of
this transformation follows the Johnson-Mehl-Avrami equation.
Transformation kinetics is slower for samples grown at low current
density and high pH. Type II deposits do not exhibit phase transfor-
mation at room temperature for a relatively long period and have
better corrosion resistance than type I ones in both acidic sodium
sulfate/borate and acidic sodium sulfate solutions.
Electrolyte pH affects the microstructure of the coatings. An in-
creasing ͑002͒ preferential orientation is observed at low pH, where
bright films are obtained. pH values in the range from 2 to 6 do not
obviously affect current efficiency or corrosion resistance of the
coatings.
7. W. E. Bradt and H. H. Oaks, Trans. Electrochem. Soc., 71, 279 ͑1937͒.
8. W. E. Bradt and L. R. Taylor, Trans. Electrochem. Soc., 73, 327 ͑1938͒.
9. C. G. Fink and M. Kolodney, Trans. Electrochem. Soc., 71, 287 ͑1937͒.
10. S. M. Shelton and M. B. Royer, Trans. Electrochem. Soc., 76, 447 ͑1939͒.
11. N. W. Ashcroft and N. D. Mermin, Solid State Physics, p. 115, W. B. Saunders,
Philadelphia, PA ͑1976͒.
12. E. V. Potter, H. C. Lukens, and R. W. Huber, Trans. AIME, 185, 399 ͑1949͒.
13. E. V. Potter and R. W. Huber, Phys. Rev., 68, 24 ͑1945͒.
14. J. E. Lewis, P. H. Scaife, and D. A. J. Swinkels, J. Appl. Electrochem., 6, 199
͑1976͒.
15. J. E. Lewis, P. H. Scaife, and D. A. J. Swinkels, J. Appl. Electochem., 6, 453
͑1976͒.
16. G. Parissis and R. Winand, in Proceedings of the Symposium on Chloride Electro-
metallurgy, The 111th AIME Meeting, Dallas, TX, 1 ͑1982͒.
17. C. L. Mantell and G. Ferment, Trans. Metall. Soc. AIME, 236, 718 ͑1966͒.
18. M. Gonsalves and D. Pletcher, J. Electroanal. Chem., 285, 185 ͑1990͒.
19. B. E. Conway and B. V. Tilak, in Advances in Catalysis, Vol. 38, D. D. Eley, Editor,
p. 58, Academic Press, New York ͑1992͒.
Acknowledgments
20. R. D. W. Kemmitt, in Comprehensive Inorganic Chemistry, Vol. 3, J. C. Bailor, Jr.,
´
H. J. Emeleus, R. Nyholm, and A. F. Trotman-Dickenson, Editors, p. 866, Perga-
This work was supported by DOD-SERDP under contract
DACA72-99-P-0201. The use of MRSEC shared facilities was sup-
ported by NSF-DMR 9809423. The authors thank an anonymous
reviewer for numerous suggestions and insightful comments.
mon Press, Oxford ͑1973͒.
21. K. Koeber and D. Schneider, Gmelin Handlook of Inorganic Chemistry, Mn, D3,
Coordination Compounds, 8th ed, Chap. 12, p. 1, Springer-Verlag Press, Berlin
͑1982͒.
22. J. Lipkowski and Z. Galus, J. Electroanal. Chem. Interfacial Electrochem., 48, 337
͑1973͒.
References
23. M. Chigane and M. Ishikawa, J. Electrochem. Soc., 147, 2246 ͑2000͒.
¨
24. E. Raub and K. Muller, Fundamentals of Metal Deposition, p. 129, Elsevier Pub-
1. A. Brenner, Electrodeposition of Alloys, Vol. I, p. 137, Academic Press, New York
͑1963͒.
lishing Company, Amsterdam ͑1967͒.
25. J. W. Christian, The Theory of Transformations in Metals and Alloys, 2nd ed.,
Chap. 1 and 12, Pergamon Press, New York ͑1975͒.
2. G. D. Wilcox and B. Petersen, Trans. IMF, 74„4…, 115 ͑1996͒.
3. M. V. Ananth, Trans. IMF, 75„6…, 224 ͑1997͒.
26. I. V. Gamali and A. V. Vorozhko, Sov. Electrochem., 16, 1011 ͑1980͒.
27. C. S. Barrett and T. B. Massalski, Structure of Metals-Crystallographic Methods,
Principles and Data, 3rd ed., p. 204, Pergamon Press, New York ͑1980͒.
28. JCPDS-International Centre for Diffraction Data, Ref. No. 17-910.
29. G. B. Harris, Philos. Mag., 43, 113 ͑1952͒.
4. S. Sriveeraraghavan, N. V. Parthasaradhy, R. M. Krishnan, S. R. Natarajan, and H.
V. K. Udupa, Met. Finish., 79„11…, 57 ͑1981͒.
5. R. S. Dean, Electrolytic Manganese and Its Alloys, Chap. 1, Reinhold Press, New
York ͑1952͒.
6. H. H. Oaks and W. E. Bradt, Trans. Electrochem. Soc., 69, 567 ͑1936͒.
Downloaded on 2014-10-06 to IP 131.215.225.9 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).