Electrodeposited Ni1-xCox nanocrystalline thin films

Article abstract of DOI:10.1149/1.2353792

Electrodeposition of nanocrystalline Ni and Ni1-x Cox thin films from chloride baths was systematically investigated by varying the electrodeposition parameters including electrolyte composition (i.e., Co2+ ion concentration), additive (i.e., saccharin), solution pH, and current density. Their effects on the film growth mechanism, film composition, residual stress, microstructure, grain size, and surface morphology were studied. Ni1-x Cox thin films electrodeposited from the baths without the addition of saccharin always showed tensile stress mode (145-367 MPa) with varying Co2+ ion concentration, solution pH, and current density. In the presence of saccharin, the Ni1-x Cox thin films showed either tensile stress or compressive stress mode, depending on the electrodeposition conditions. Especially, it was observed from a cross-sectional TEM observation that Ni thin film electrodeposited from the bath containing saccharin exhibited the formation of an amorphous Ni layer (about 300 nm thick) at the initial stage of the film growth. Also, Ni1-x Cox thin films electrodeposited from the bath with/without the addition of saccharin showed the formation of the interface phase layer (about 10-110 nm thick), which has the chemical composition of 50 atom % Ni and 50 atom % Co.The Electrochemical Society

Full text of DOI:10.1149/1.2353792

C814
Journal of The Electrochemical Society, 153 ͑12͒ C814-C821 ͑2006͒
0013-4651/2006/153͑12͒/C814/8/$20.00 © The Electrochemical Society
Electrodeposited Ni1−xCox Nanocrystalline Thin Films
Structure–Property Relationships
,z
D.-Y. Park,a, K. S. Park,a J. M. Ko,b D.-H. Cho,c S. H. Lim,c W. Y. Kim,d
*
B. Y. Yoo,e and N. V. Myunge,
*
aDepartment of Applied Materials Engineering and bDepartment of Chemical Technology,
Hanbat National University, Daejeon 305-719, Korea
cDepartment of Advanced Materials Science and Engineering, Kangwon National University, Chuncheon,
Kangwon-do 200-701, Korea
dAdvanced Materials Research and Development Center, Korea Institute of Industrial Technology,
Incheon 406-130, Korea
eDepartment of Chemical and Environmental Engineering, University of California at Riverside, Riverside,
California 92521, USA
Electrodeposition of nanocrystalline Ni and Ni1−xCox thin films from chloride baths was systematically investigated by varying the
electrodeposition parameters including electrolyte composition ͑i.e., Co2+ ion concentration͒, additive ͑i.e., saccharin͒, solution
pH, and current density. Their effects on the film growth mechanism, film composition, residual stress, microstructure, grain size,
and surface morphology were studied. Ni1−xCox thin films electrodeposited from the baths without the addition of saccharin always
showed tensile stress mode ͑145–367 MPa͒ with varying Co2+ ion concentration, solution pH, and current density. In the presence
of saccharin, the Ni1−xCox thin films showed either tensile stress or compressive stress mode, depending on the electrodeposition
conditions. Especially, it was observed from a cross-sectional TEM observation that Ni thin film electrodeposited from the bath
containing saccharin exhibited the formation of an amorphous Ni layer ͑about 300 nm thick͒ at the initial stage of the film growth.
Also, Ni1−xCox thin films electrodeposited from the bath with/without the addition of saccharin showed the formation of the
interface phase layer ͑about 10–110 nm thick͒, which has the chemical composition of 50 atom % Ni and 50 atom % Co.
© 2006 The Electrochemical Society. ͓DOI: 10.1149/1.2353792͔ All rights reserved.
Manuscript submitted March 7, 2006; revised manuscript received August 2, 2006. Available electronically October 9, 2006.
Electrodeposited Ni and Ni1−xCox thin films have been inten-
ited Ni film containing greater than 170 ppm sulfur was highly
notch sensitive. Film hardness was shown to be a direct function of
sulfur content.
sively studied because of their potential applications in microelec-
tromechanical systems ͑MEMS͒,1-3 LIGA,4 electronic devices ͑e.g.,
spintronics, magnetic storages, optical recording devices, and inte-
grated sensors͒,5,6 and several other industries ͑e.g., automotive,
manufacturing, and chemical process͒.7 Ni and Ni1−xCox thin/thick
films have been incorporated into magnetic-MEMS devices includ-
ing sensors, microactuators, micromotors, and frictionless microge-
ars because of their excellent physical properties.1,3,4,8 Magnetic
layer thickness in MEMS can range from a few nanometers to hun-
dreds of micrometers, depending on the applications. Residual stress
of the electrodeposited films is an important factor for incorporating
magnetic materials in MEMS devices. High stress in magnetic thin/
thick films may result in the malfunctioning of MEMS devices be-
cause of delamination of the deposited films from MEMS structures.
Numerous studies of the electrodeposited Ni thin/thick films
have been carried out.4,9-11 Typical nickel plating baths are chloride,
sulfate, sulfamate, and mixed sulfate-chloride ͑i.e., Watts-type͒
baths.11 When a sufficient amount of sulfur-containing additives
such as benzoyl sulfamide ͑saccharin͒ or naphthalene trisulfonic
acid ͑NTSA͒ is added to Ni plating baths, residual stress of Ni
electrodeposits can be reduced to almost none. When added in
greater quantities, these additives result in compressively stressed Ni
deposits. Weil12-14 reviewed various methods to measure residual
stress of the electrodeposited films, the effects of electrodeposition
parameters on residual stress in various pure metals ͑e.g., Cu, Ni,
Co, and Zn͒, and the suggested hypotheses for the causes of residual
stress in the electrodeposited films. Even though several hypotheses
have been proposed to explain the causes of residual stress in the
electrodeposited films, no overall hypothesis has been formulated to
date. Dini15 also reviewed residual stress in various electrodeposited
materials. Ni electrodeposited from sulfamate baths exhibited the
lowest stress value of 59 MPa compared to sulfate, chloride, bro-
mide, and fluoroborate baths. Dini et al.11 investigated the influence
of sulfur on the properties of the electrodeposited Ni using a Charpy
test and Knoop hardness test. They reported that the electrodepos-
Even though there are many studies on the electrodeposited
Ni1−xCox thin films, they focused mostly on anomalous
codeposition,16-20 the applications of the magnetic Ni1−xCox materi-
als in MEMS devices,3 mechanical properties,21 electrochemical
properties,22,23 or anion effects.24 There is a lack of detailed studies
correlating the electrodeposition parameter and the role of sulfur-
containing additive ͑e.g., saccharin͒ with the property in the elec-
trodeposited Ni1−xCox thin films. Especially, binary iron group al-
loys including NiCo, NiFe, and FeCo exhibit a so-called anomalous
codeposition in which the less noble metal is deposited preferen-
tially under certain conditions.16-20 The mechanism of anomalous
codeposition in iron group alloys was studied by several researchers.
The behavior of anomalous codeposition may be attributed to an
inhibition effect of the less noble metal on the more noble metal.
This paper is the extended research work of our previous result.25
In this study, we investigated the effect of the electrodeposition pa-
rameters including metal ion concentration, solution pH, current
density, and saccharin on film growth mechanism, residual stress,
microstructure, grain size, and surface morphology of the electrode-
posited Ni and Ni1−xCox thin films. Specifically, We studied the
effect of an impurity element ͑sulfur͒ on the growth mechanism of
Ni and Ni1−xCox thin films using a cross-sectional transmission elec-
tron microscopy ͑TEM͒ observation.
Experimental
Ni and Ni1−xCox thin films were electrodeposited from chloride
baths. Table I lists the plating solution composition investigated.
NaCl and boric acid were used as supporting electrolyte and pH
buffer, respectively. Solution was exposed to air and solution pH
was adjusted with KOH or HC1. Ni and Ni1−xCox thin films were
galvanostatically electrodeposited using an EG&G PAR
potentiostat/galvanostat ͑model 263A or 273͒ at room temperature
without stirring. The cathode current efficiency was deduced from
differential mass measurement. The effects of the solution
Co2+/Ni2+ ratio on the properties including film composition, cur-
rent efficiency, residual stress, microstructure and surface morphol-
ogy of Ni1−xCox films were studied by varying Co2+ concentration
*
Electrochemical Society Active Member.
z E-mail: dypark@hanbat.ac.kr
Downloaded on 2015-06-19 to IP 128.122.253.228 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, 153 ͑12͒ C814-C821 ͑2006͒
C815
Table I. Bath compositions and operating conditions (unless oth-
erwise noted) for Ni and Ni1−xCox thin films electrodeposited
from chloride baths „M = mol dm−3….
Chemical/condition
Concentration ͑M͒
Ni2+ ͑as NiCl2·6H2O͒
Co2+ ͑as CoCl2·6H2O͒
NaCl
0.2
0–0.206
0.7
H3BO3
Saccharin
0.4
0 or 0.01
pH
2–6
Current density
Temperature
1–25 mA cm−2
Room temperature ͑24°C͒
from 0.03 to 0.206 M, with 0.2 M Ni2+ ion concentration, solution
pH 4, and 10 mA cm−2. The effects of solution pH on the properties
of Ni1−xCox films were investigated by varying solution pH from 2
to 6, with 0.2 M Ni2+ and 0.05 M Co2+ ion concentrations, and
10 mA cm−2. Also, the effects of current density ͑CD͒ on the prop-
erties of Ni1−xCox films were investigated by varying current density
from 1 to 25 mA cm−2, with 0.2 M Ni2+ and 0.05 M Co2+ ion con-
centrations, and solution pH 4. Ni and Ni1−xCox films were elec-
trodeposited on Cu substrates ͑PN 1194, Specialty Testing & Devel-
opment Co.͒; nickel sheet was used as a soluble anode. The residual
stress of the Ni1−xCox films was measured by a deflection method
with a deposit stress analyzer ͑model 683, Specialty Testing & De-
velopment Co.͒ and calculated by Stoney’s equation.15,26,27 Figure
1a shows a schematic of the apparatus for the electrodeposition of
Ni and Ni1−xCox thin films to measure residual stress of the thin
films. A Cu test strip ͑PN 1194, Specialty Testing & Development
Co.͒ with a constant electrodeposited area of 7.74 cm2 is located at
the center of the storage tank ͑a total volume of 2.5 L͒. Figure 1b
and c shows the dimension of the Cu test strip for measuring re-
sidual stress of the thin film. One side of the leg in the Cu test strip
was laminated with polymer to prevent the electrodeposition of Ni
or Ni1−xCox films. The other side ͑cross-hatched area͒ of the leg was
not laminated with polymer to deposit the thin film. Electrodeposi-
tion of Ni or Ni1−xCox films was conducted without stirring of elec-
trolyte. Stirring of electrolyte can affect the stress value because the
Cu test strip is very thin ͑0.06 mm thick͒ and flexible. Cross-
sectional TEM ͑model JEM-2010, JEOL, Ltd.͒ was used to charac-
terize the microstructure of Ni and Ni1−xCox films. Structural infor-
mation was obtained using electron diffraction and a bright-field
image, operating at 200 kV. Samples for TEM observation were
prepared by a conventional procedure, including mechanical polish-
ing, dimpling, and Ar ion milling or focused ion beam milling ͑FIB͒.
Surface morphology and film composition were examined using
scanning electron microscopy ͑SEM͒ ͑model JSM-6300, JEOL,
Ltd.͒ operating at 20 kV and the working distance of 17 mm, and
energy-dispersive spectroscopy ͑EDS͒ ͑model ISIS, Oxford Instru-
ments͒, respectively. An X-ray diffractometer ͑XRD͒ ͑model
D/MAX 2500H, RIGAKU͒ with Cu K␣ radiation ͑operating at
45 kV͒ was used for the identification of the phase and the measure-
ment of the grain size in the thin films. The conditions of XRD were
a scanning range of 30–100° with 0.03° increments and a 1 s col-
lection time per increment.
Figure 1. ͑a͒ Schematic of apparatus for electrodeposition, ͑b͒ dimensions of
Cu test strip for measuring residual stress of thin film, cross-hatched area is
electrodeposited by Ni or CoNi, and ͑c͒ deflection of Cu test strip after
electrodeposition ͑␦: relative displacement͒. Unit: mm.
Ni ͑or close to amorphous Ni͒. In the absence of saccharin, the
formation of the amorphous layer at the early stage of electrodepo-
sition was not observed ͑Fig. 2a͒. The amount of sulfur from EDS
analysis was measured to be about 1 atom % in Ni thin film ele-
crodeposited from the bath containing 0.01 M saccharin. It is be-
lieved that the impurity element of sulfur from the source of saccha-
rin influenced the film growth mechanism in Ni thin film. Figures 3
and 4 show the cross-sectional TEM micrographs of Ni1−xCox thin
films electrodeposited with/without the addition of saccharin. The
Ni1−xCox thin films were electrodeposited from the bath containing
0.2 M NiCl2 and 0.05 M CoCl2 at 10 mA cm−2, pH 4.0, and room
temperature without stirring. The deposit thickness of Ni1−xCox thin
films was measured to be about 2.7 ␮m from the cross-sectional
TEM micrographs. The Ni1−xCox thin film electrodeposited without
the addition of saccharin ͑Fig. 3͒ exhibits the formation of a thicker
and continuous interface phase layer ͑about 110 nm thick͒ at the
beginning of the film growth along the interface between the Cu
substrate and Ni1−xCox thin films. The chemical composition of the
interface phase layer ͑point A in Fig. 3a͒ was measured to be
49 atom % Ni and 51 atom % Co from EDS analysis as shown in
Fig. 5. The chemical composition of the remaining film ͑point B, C,
and D in Fig. 3a͒ in Ni1−xCox thin film was 71 atom % Ni and
29 atom % Co shown in Fig. 6. EDS analysis for the remaining film
was conducted from the near interface phase layer to the top of the
remaining film at three different positions ͓the position near the
Results and Discussion
Figure 2 shows the cross-sectional TEM micrographs of Ni thin
films electrodeposited with/without the addition of saccharin. In the
presence of saccharin, an amorphous layer ͑approximately 300 nm
thick͒ grows first on the Cu substrate at the beginning of elec-
trodeposition and then the nanocrystalline Ni layer ͑or close to
amorphous Ni͒ is electrodeposited ͑Fig. 2b͒ as shown in the bright-
field image and the corresponding selected area diffraction pattern
͑SADP͒. The SADP clearly shows that the first layer from the Cu
substrate is amorphous Ni and the second Ni layer nanocrystalline
Downloaded on 2015-06-19 to IP 128.122.253.228 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C816
Journal of The Electrochemical Society, 153 ͑12͒ C814-C821 ͑2006͒
Figure 2. Cross-sectional TEM bright field images and SAD patterns of Ni
thin films electrodeposited with/without saccharin: ͑a͒ no saccharin and ͑b͒
0.01 M saccharin. Deposit thickness approximately 3 ␮m.
Figure 3. Cross-sectional TEM bright-field images and SAD pattern of
Ni1−xCox thin films electrodeposited without saccharin: ͑a͒ low magnification
and ͑b͒ high magnification. Deposit thickness approximately 3 ␮m.
interface phase ͑Fig. 6a͒, the intermediate position ͑Fig. 6b͒, and the
near top surface position ͑Fig. 6c͒ of the remaining film͔ to check on
any compositional differences between the remaining film. The
chemical compositions were measured to be nearly same at the three
different areas. The peak of the chemical element Fe in EDS analy-
sis resulted from the Cu substrate itself. The amount of Fe element
was found to be higher at the area near to the Cu substrate compared
to the top surface area. Fe was not detected at the near top surface
position ͑point D in Fig. 3a͒ of the Ni1−xCox thin films, as shown in
Fig. 6c. Figure 4 shows that the Ni1−xCox thin films electrodeposited
with the addition of saccharin also exhibited a thinner and discon-
tinuous interface phase layer ͑less than 20 nm thick͒. The chemical
composition of the remaining film was also measured to be
70 atom % Ni and 30 atom % Co from EDS analysis. Therefore, it
is evident that the addition of element Co in Ni thin film results in
the formation of the interface phase layer, which consists of
50 atom % Ni and 50 atom % Co, at the early stage of electrodepo-
sition. The addition of saccharin in Ni1−xCox thin films results in the
formation of a thinner and discontinuous interface phase layer. SAD
patterns from the interface phase layer could not be taken because
the aperture size in TEM compared to thickness of the interface
phase layer is too large to get SAD patterns. Further investigation to
reveal the exact growth mechanism at the beginning stage of elec-
trodeposition in Ni and Ni1−xCox thin films is needed.
ing current density from 1 to 25 mA cm−2; the deposit Co content
decreased from 72 to 22 atom %, as shown in Fig. 7b. Current den-
sity strongly influenced the Ni2+/Co2+ ratio on the electrodeposited
Ni1−xCox thin films, but the Ni2+/Co2+ ratio was independent of the
addition of saccharin. Zech et al.18-20 experimentally and theoreti-
cally studied anomalous codeposition of iron group alloys. They
reported that Ni deposition was inhibited by codeposition of Co.
They suggested that inhibition or enhancement of the deposition rate
is attributed to interactions between codepositing species. Therefore,
the higher Co content ͑72–60 atom % at low current density ͑1 to
3 mA cm−2͒ in Ni1−xCox thin films electrodeposited from the baths
with/without the addition of saccharin may result from the inhibition
of Ni by codeposition of Co. From our previous paper,25 the depos-
ited Co content increased from 23 to 85 atom % with increasing
Co2+ ion concentration in the plating baths from 0.03 to 0.206 M.
The Ni2+/Co2+ ratio on the electrodeposited Ni1−xCox thin films was
nearly identical in the absence or presence of saccharin. Deposited
sulfur content in Ni1−xCox thin films electrodeposited from the baths
containing 0.01 M saccharin were measured to be less than
1 atom % and was independent of Co2+ ion concentration. The de-
tection limit of EDS for chemical elements may be about 1%. The
sulfur amount of less than 1 atom % in this study is very small and
may be in the range of detection limit. However, if we consider the
effect of sulfur on film stress, microstructure, and growth mecha-
nism in Ni1−xCox thin film, the small amount of sulfur is not negli-
gible.
Figure 7 shows the change in the film compositions as a function
of solution pH, and current density in the electrolyte solution and in
the presence or absence of saccharin. Figure 7a shows that the de-
posit Ni and Co contents with/without saccharin are independent of
solution pH. The Ni2+/Co2+ ratio on the electrodeposited Ni1−xCox
thin films was independent of the addition of saccharin. However,
the deposit Ni content increased from 28 to 78 atom % with increas-
Figure 8 shows the dependence of current efficiency on Co2+ ion
concentration ͑Fig. 8a͒, solution pH ͑Fig. 8b͒, and current density
Downloaded on 2015-06-19 to IP 128.122.253.228 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, 153 ͑12͒ C814-C821 ͑2006͒
C817
Figure 6. ͑Color online͒ EDS spectra showing the same chemical composi-
tion ͑71 atom % Ni–29 atom % Co͒ of nanocrystalline Ni1−xCox thin films
electrodeposited without saccharin at ͑a͒ near interface phase ͑point B in Fig.
3a͒, ͑b͒ intermediate area ͑point C in Fig. 3a͒, and ͑c͒ near top surface area
͑point D in Fig. 3a͒.
Figure 4. Cross-sectional TEM bright-field images and SAD pattern of
Ni1−xCox thin films electrodeposited with 0.01 M saccharin: ͑a͒ low magni-
fication and ͑b͒ high magnification. Deposit thickness approximately 3 ␮m.
that about 2.7 ␮m thick films were obtained despite using the depo-
sition time for obtaining 3 ␮m thick films, as shown in Fig. 3a and
4a. There is good agreement between the current efficiency and the
deposit film thickness in this study. Fenineche et al.21 reported that
the high current efficiency ͑93%͒ in CoNi electrodeposits from chlo-
ride baths was obtained at room temperature, pH 3.5, and current
densities from 10 to 15 mA cm−2. High current efficiency of more
than 90% in this study appears to be in good agreement with the
reported value ͑93%͒ by Fenineche et al.21 Golodnitsky et al.22 stud-
ied NiCo alloy electrodeposited from a sulfamate electrolyte with
different anion additives. They observed an increase in the current
efficiency from 65 to 82%, with increasing current density from 2 to
100 mA/cm2. Solution pH and operating temperature were kept at 3
and 55°C, respectively. Therefore, it is believed that electrodeposi-
tion conditions including anion type, current density, and operating
temperature strongly influenced the current efficiency.
͑Fig. 8c͒ in Ni and Ni1−xCox thin films. Current efficiency of more
than about 90% was almost independent of Co2+ ion concentration,
whereas current efficiency slightly increased with increasing solu-
tion pH and slightly decreased with increasing current density. From
the cross-sectional TEM analysis in this study, it was also observed
Figure 9 shows the change of residual stress in electrodeposited
Ni1−xCox alloys as a function of solution pH and current density in
the plating solution with/without the addition of saccharin. Residual
stresses of Ni1−xCox thin films with/without the addition of saccharin
show a minimum at solution pH 4 ͑Fig. 9a͒. Unfortunately, previous
data of solution pH effect on residual stress in Ni1−xCox thin films is
not available for comparison. Weil13 has summarized that there is a
Figure 5. ͑Color online͒ EDS spectra showing chemical composition of the
interface phase layer ͑point A in Fig. 3a͒ in Ni1−xCox thin films electrode-
posited without saccharin; 49 atom % Ni–51 atom % Co.
Downloaded on 2015-06-19 to IP 128.122.253.228 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C818
Journal of The Electrochemical Society, 153 ͑12͒ C814-C821 ͑2006͒
Figure 7. Dependence of film composition on solution pH and current den-
sity in Ni1−xCox thin films: ͑᭝, ᮀ ͒ no saccharin and ͑᭿, ᭡ , b ͒ 0.01 M
saccharin. Deposit thickness about 3 ␮m.
sharp rise of residual stress in Ni deposits in the range of solution
pH values from 4 to 6. The exact value of solution pH where the
stress begins to increase appears to depend greatly on plating con-
ditions and the bath compositions. Ni1−xCox thin films in this study
exhibited a little higher value of residual stress above solution pH 5.
Residual stress of Ni1−xCox film from saccharin-free electrolytes
was independent of current density ͑approximately 220 MPa͒ at low
current density ͑1–10 mA cm−2͒. However, Ni1−xCox films from
saccharin-free electrolytes exhibited higher residual stress values of
approximately 360 MPa at 15 and 20 mA cm−2. Ni1−xCox electrode-
posits from the baths containing 0.01 M saccharin show minimum
residual stress of approximately 0 MPa at 10 mA cm−2 ͑Fig. 9b͒.
From our previous paper,25 we replotted the stress changes in Ni and
Ni1−xCox thin films as a function of deposit Co content ͑Fig. 9c͒.
Residual stress of 145 MPa ͑tensile stress mode͒ was obtained in Ni
thin film electrodeposited from saccharin-free electrolyte. Ni thin
film deposited from saccharin-containing electrolyte exhibited a
compressive stress mode ͑−85 MPa͒. It has been reported that Ni
electrodeposited from additive-free Watts baths exhibits tensile
stresses of 125–185 MPa.28 Kushner29 reported tensile stress values
of 59–228 MPa from the baths with various different anions. They
reported that the residual stress of Ni electrodeposited from chloride
bath was measured to be 228 MPa. The difference between the re-
ported value of residual stress ͑228 MPa͒ and that of this study
͑145 MPa͒ may be attributed to the different substrate and elec-
trodeposition conditions, including Ni2+ ion concentration and cur-
rent density. It was suggested that codeposition of elemental sulfur
Figure 8. Dependence of current efficiencies in Ni1−xCox thin films on ͑a͒
Co2+ ion concentration, ͑b͒ solution pH, and ͑c͒ current density; ͑᭺͒ no
saccharin and ͑᭝͒ 0.01 M saccharin. Deposit thickness about 3 ␮m.
in Ni electrodeposited from the bath with sulfur-containing organic
additives results in the transition of residual stress from tensile to
compressive mode.13,28 As deposited Co content in Ni1−xCox thin
films increased up to about 89 atom % in the absence of saccharin,
residual stress of Ni1−xCox thin films increased from 173 to
262 MPa. Ni1−xCox thin films electrodeposited from the baths con-
taining saccharin exhibited
a
transition from compressive
Downloaded on 2015-06-19 to IP 128.122.253.228 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, 153 ͑12͒ C814-C821 ͑2006͒
C819
Figure 10. Dependence of surface morphology in Ni1−xCox thin films on
solution pH and current density with/without saccharin. Deposit thickness
approximately 3 ␮m.
Figure 9. Dependence of residual stress in Ni1−xCox thin films electrodepos-
ited from chloride baths on ͑a͒ solution pH, ͑b͒ current density, and ͑c͒
deposit Co content; ͑᭺͒ no saccharin and ͑᭝͒ 0.01 M saccharin. Deposit
thickness approximately 3 ␮m.
effectively as in Ni deposits. Thus, it can be hypothesized that the
transition of residual stress from compressive to tensile stress mode
in Ni1−xCox deposits with increasing Co2+ ion concentration may be
attributed to the lesser effect of sulfur in Co-rich deposits.
Figure 10 shows the dependence of surface morphology in
Ni1−xCox thin films with/without saccharin on solution pH ͑Fig.
10a-d͒, and current density ͑Fig. 10e-h͒. It is evident that the addi-
tion of 0.01 M saccharin in Ni1−xCox plating baths strongly influ-
enced the surface morphology. The Ni1−xCox thin films, which con-
sist of about 35 atom % Co and about 65 atom % Ni in the absence
of saccharin, exhibited a nodular surface morphology ͑Fig. 10a and
c͒ with increasing solution pH from 2 to 6. In the presence of sac-
͑−51 MPa͒ to tensile stress mode ͑120 MPa͒. Almost zero-stress
Ni1−xCox thin film was obtained at Co deposit content of
30–40 atom %. Small quantities ͑0.01–0.1 g/L͒ of most sulfur-
containing additives in the plating bath rapidly reduce residual stress
in Ni deposits.15 It is well known that the saccharin as a source of
sulfur effectively decreases the residual stress in Ni deposits.13 The
addition of saccharin lowers residual stress in Co deposits, but not as
Downloaded on 2015-06-19 to IP 128.122.253.228 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C820
Journal of The Electrochemical Society, 153 ͑12͒ C814-C821 ͑2006͒
Figure 11. XRD patterns of Ni1−xCox
thin films electrodeposited from plating
bath with/without saccharin as a function
of solution pH, and current density. Ni2+
:
0.2 M, Co2+: 0.05 M. Deposit thickness
approximately 3 ␮m ͑S: substrate͒.
the Scherrer formula. The average grain sizes of fcc ͑111͒ and fcc
͑200͒ in Ni1−xCox thin films ͑Fig. 11b͒ electrodeposited from the
baths containing 0.01 M saccharin were measured to be 29 and
16 nm, respectively. The average grain sizes of fcc ͑111͒ and fcc
͑200͒ in Ni1−xCox thin films ͑Fig. 11c͒ electrodeposited from
saccharin-free electrolyte were measured to be 49 and 44 nm, re-
spectively. Also, the average grain sizes of fcc ͑111͒ and fcc ͑200͒ in
Ni1−xCox thin films ͑Fig. 11d͒ electrodeposited from the baths con-
taining 0.01 M saccharin were measured to be 32 and 16 nm, re-
spectively. Therefore, it is evident that the addition of saccharin
strongly influenced the grain size of the dominant phases. There
exists a good agreement in grain refinement due to the addition of
saccharin in Ni and Ni1−xCox thin films from XRD analysis and
SAD patterns ͑Fig. 2b and Fig. 4a͒ in TEM. It was suggested that
the stress difference between Ni1−xCox thin films with and without
0.01 M saccharin may result from the addition of saccharin in the
plating baths ͑as a source of impurity element, sulfur͒ rather than the
microstructure.25 The amount of sulfur from EDS analysis was mea-
sured to be less than 1 atom % in Ni and Ni1−xCox thin films elec-
trodeposited from the bath containing 0.01 M saccharin. The thin
films deposited by physical or chemical vapor deposition or elec-
trodeposition exhibit the residual stress, resulting in cracking or
peeling the film.30 If the pure thin film was deposited at high tem-
perature and operated at lower temperature, the residual stress in
thin film may be attributed to a mismatch in the coefficient of ther-
mal expansion between the substrate and the film. If the pure thin
film was deposited on a substrate under stress, the residual stress
may result from a mismatch in the elastic modulus. It was also
suggested that impurities and defects influence the residual stress of
the thin film during deposition or solid-state transformation.30,31 The
thin film containing residual gas, oxygen, and water vapor usually
exhibit compressive stress mode.32,33 Therefore, it could be summa-
charin, the Ni1−xCox thin films showed the smooth surface morphol-
ogy ͑Fig. 10b and d͒. At low current density of 1 mA/cm2, the
Ni1−xCox thin film that consists of about 75 atom % Co and about
25 atom % Ni exhibited a pyramidal surface morphology ͑Fig. 10e͒.
At high current density, the Ni1−xCoxthin film that consists of about
20 atom % Co and about 80 atom % Ni showed a nodular surface
morphology ͑Fig. 10g͒.
Figure 11 shows the dependence of the XRD patterns in
Ni1−xCox thin films with/without the addition of saccharin on solu-
tion pH ͑Fig. 11a and b͒, and current density ͑Fig. 11c and d͒. The
XRD patterns of Ni1−xCox thin films electrodeposited from
saccharin-free electrolyte consist of face-centered cubic ͑fcc͒ ͑111͒,
fcc ͑200͒, fcc ͑220͒ ͓or hexagonal close-packed ͑hcp͒ ͑110͔͒,
fcc͑311͒ ͓or hcp ͑112͔͒, and fcc ͑222͒ at solution pH 2–6, as shown
in Fig. 11a. Ni1−xCox thin films deposited from the baths containing
0.0.1 M saccharin ͑Fig. 11b͒ exhibited a similar microstructure with
different intensity compared to Ni1−xCox thin films electrodeposited
from saccharin-free electrolyte in Fig. 11a. Ni1−xCox thin films elec-
trodeposited from saccharin-free electrolyte exhibited mixed hcp
and fcc phases at 1 mA cm−2; fcc ͑220͒ or hcp ͑110͒ is a dominant
phase. The dominant phases in Ni1−xCox thin film electrodeposited
at 5 and 10 mA cm−2 are fcc ͑111͒ and fcc ͑200͒. The dominant
phases are changed to be fcc ͑220͒ or hcp ͑110͒ at 20 and
25 mA cm−2. The dominant phases are fcc ͑111͒ and fcc ͑200͒ in
Ni1−xCox thin films deposited from the baths containing 0.01 M sac-
charin. There was not found any trend in microstructure with in-
creasing current density. In the presence of saccharin, the line broad-
ening of the XRD peaks of the dominant phases ͓fcc ͑111͒, fcc
͑200͒, and hcp ͑112͒ or fcc ͑311͔͒ was clearly observed. The average
grain sizes of fcc ͑111͒ and fcc ͑200͒ in Ni1−xCox thin films ͑Fig.
11a͒ electrodeposited from saccharin-free electrolyte were measured
to be 55 and 48 nm, respectively. The grain size was calculated from
Downloaded on 2015-06-19 to IP 128.122.253.228 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, 153 ͑12͒ C814-C821 ͑2006͒
C821
rized that the purer the film, the greater the tendency for tension, and
References
oxygen and other impurities tend to give rise to compressive stress
1. J. W. Judy and R. S. Muller, J. Microelectromech. Syst., 6, 249 ͑1997͒.
2. K. Kataoka, S. Kawamura, T. Itoh, K. Ishikawa, H. Honma, and T. Suga, Sens.
Actuators, A, 103, 116 ͑2003͒.
3. H. H. Yang, N. V. Myung, J. Yee, D.-Y. Park, B.-Y. Yoo, M. Schwartz, K. Nobe,
and J. W. Judy, Sens. Actuators, A, 97-98, 88 ͑2002͒.
mode.34
Conclusion
Residual stresses of Ni1−xCox thin films were strongly influenced
by current density and Co2+ ion concentrations. Solution pH slightly
affected residual stress. The addition of saccharin in the Ni plating
bath resulted in the changes of film growth mechanism, residual
stress, surface morphology, grain size, and microstructure. The ad-
dition of saccharin in Ni thin film resulted in the formation of an
amorphous Ni layer about 300 nm thick at the initial stage of film
growth. The continuous interface phase layer ͑110 nm thick͒ which
has different composition of 50 atom % Ni and 50 atom % Co was
formed in Ni1−xCox thin films at the initial stage of film growth. The
addition of saccharin in Ni1−xCox thin films resulted in not only the
decrease of the thickness of the interface phase layer ͑from 110 to
50 nm͒ but also the formation of a discontinuous interface phase
layer ͑less than 20 nm thick͒. For saccharin-free electrolyte, Ni thin
film showed tensile stress mode ͑145 MPa͒. However, Ni thin film
electrodeposited from the bath containing 0.01 M saccharin exhib-
ited compressive stress mode ͑−85 MPa͒. The transition of residual
stress from compressive to tensile stress mode in Ni1−xCox thin films
electrodeposited from the baths containing 0.01 M saccharin was
observed with increasing Co2+ ion concentrations. It was suggested
that the effect of saccharin ͑as a source of an impurity element,
sulfur͒ on residual stress was less effective in Co-rich thin films
compared to Ni-rich thin films. Therefore, Co-rich thin films show a
tensile stress mode despite the addition of saccharin in the plating
bath.
4. S. M. Allameh, J. Lou, F. Kavishe, T. Buchheit, and W. O. Soboyejo, Mater. Sci.
Eng., A, 371, 256 ͑2004͒.
5. F. Czerwinski, A. Zielinska-Lipiec, and J. A. Szpunar, Acta Mater., 47, 2553
͑1999͒.
6. C. A. Moina and M. Vazdar, Electrochem. Commun., 3, 159 ͑2001͒.
7. C.-C. Hu, C.-Y. Lin, and T.-C. Wen, Mater. Chem. Phys., 44, 233 ͑1996͒.
8. J. W. Judy, R. S. Muller, and H. H. Zappe, J. Microelectromech. Syst., 4, 162
͑1995͒.
9. W. M. Yin, S. H. Whang, and R. A. Mirshams, Acta Mater., 53, 383 ͑2005͒.
10. K. P. Larsen, A. A. Rasmussen, J. T. Ravnkilde, M. Ginnerup, and O. Hansen, Sens.
Actuators, A, 103, 156 ͑2003͒.
11. J. W. Dini, H. R. Johnson, and H. J. Saxton, J. Vac. Sci. Technol., 12, 766 ͑1975͒.
12. R. Weil, Plating, 57, 1231 ͑1970͒.
13. R. Weil, Plating, 58, 50 ͑1971͒.
14. R. Weil, Plating, 58, 137 ͑1971͒.
15. J. W. Dini, Electrodeposition-The Materials Science of Coatings and Substrates,
Chap. 9, Noyes Publications, Park Ridge, NJ ͑1993͒.
16. S. Hessami and C. W. Tobias, J. Electrochem. Soc., 136, 3611 ͑1989͒.
17. K. Y. Sasaki and J. B. Talbot, J. Electrochem. Soc., 142, 775 ͑1995͒.
18. N. Zech, E. J. Podlaha, and D. Landolt, J. Appl. Electrochem., 28, 1251 ͑1998͒.
19. N. Zech, E. J. Podlaha, and D. Landolt, J. Electrochem. Soc., 146, 2886 ͑1999͒.
20. N. Zech, E. J. Podlaha, and D. Landolt, J. Electrochem. Soc., 146, 2892 ͑1999͒.
21. N. Fenineche, C. Coddet, and A. Saida, Surf. Coat. Technol., 41, 75 ͑1990͒.
22. D. Golodnitsky, N. V. Gudin, and G. A. Volyanuk, J. Electrochem. Soc., 147, 4156
͑2000͒.
23. A. Bai and C.-C. Hu, Electrochim. Acta, 47, 3447 ͑2002͒.
24. D. Golodnitsky, Yu. Rosenberg, and A. Ulus, Electrochim. Acta, 47, 2707 ͑2002͒.
25. D.-Y. Park, R. Y. Song, J. M. Ko, B. Y. Yoo, and N. V. Myung, Electrochem.
Solid-State Lett., 8, C23 ͑2005͒.
26. B.-Z. Lee and D. N. Lee, Acta Mater., 46, 3701 ͑1998͒.
27. Electronic Thin Film Science for Electrical Engineers and Materials Scientists,
K.-N. Tu, J. W. Mayer, and L. C. Feldman, Editors, Chap. 4, Macmillan Publishing
Company, New York ͑1992͒.
28. Modern Electroplating, 4th ed., G. A. Di Bari, M. Schlesinger, and M. Paunovic,
Editors, pp. 139–199, Wiley-Interscience, New York ͑2000͒.
29. J. B. Kushner, Met. Finish., 56, 81 ͑1958͒.
Acknowledgments
This work was supported by grant no. R01-2003-000-10597-0
from the Basic Research Program of the Korea Science & Engineer-
ing Foundation. B.Y.Y. and N.V.M. acknowledge funding support
from the American Electroplaters and Surface Finishers ͑AESF͒,
project no. 115.
30. F. A. Smidt, Int. Mater. Rev., 35, 61 ͑1990͒.
31. W. Esinger, Rev. Sci. Instrum., 63, 5217 ͑1992͒.
32. H. Leplan, J. Y. Robic, and Y. Pauleau, J. Appl. Phys., 79, 6926 ͑1996͒.
33. H. Leplan, B. Geenen, and J. Y. Robic, J. Appl. Phys., 78, 962 ͑1995͒.
34. S. Tamulevičius, Vacuum, 51, 127 ͑1998͒.
Hanbat National University assisted in meeting the publication costs of
this article.
Downloaded on 2015-06-19 to IP 128.122.253.228 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).