Growth kinetics of electroless cobalt deposition by TEM

Article abstract of DOI:10.1149/1.1793691

The activation energy and initial growth of electroless cobalt thin films on silicon wafer has been studied by transmission electron microscopy (TEM). Electroless cobalt was deposited on Si(001) substrates after preactivation treatment. The electroless cobalt deposited initially nucleated into particles with different sizes and shapes, which coalesced together sequentially, and finally became a continuous film. The incubation period of Co-P deposit decreased when plating temperature increased from 65 to 90°C. The plot of film thickness vs. plating time was used to determine the plating rates. The plating rates increased when plating temperature increased from 65 to 90°C. The activation energy for the initial deposition of electroless Co-P film was 15.67 kcal/mol, which was calculated according to the Arrhenius equation.

Full text of DOI:10.1149/1.1793691

C680
Journal of The Electrochemical Society, 151 ͑10͒ C680-C683 ͑2004͒
0
013-4651/2004/151͑10͒/C680/4/$7.00 © The Electrochemical Society, Inc.
Growth Kinetics of Electroless Cobalt Deposition by TEM
a
a,z
a
b
W. L. Liu, S. H. Hsieh, T. K. Tsai, and W. J. Chen
aDepartment of Materials Science and Engineering, National Huwei University of Science and Technology,
Huwei, Yunlin, Taiwan 632
b
Graduate Institute of Materials Engineering, National Pingtung University of Science and Technology,
Neipu Hsiang, Pingtung, Taiwan
The activation energy and initial growth of electroless cobalt thin films on silicon wafer has been studied by transmission electron
microscopy ͑TEM͒. Electroless cobalt was deposited on Si͑001͒ substrates after preactivation treatment. The electroless cobalt
deposited initially nucleated into particles with different sizes and shapes, which coalesced together sequentially, and finally
became a continuous film. The incubation period of Co-P deposit decreased when plating temperature increased from 65 to 90°C.
The plot of film thickness vs. plating time was used to determine the plating rates. The plating rates increased when plating
temperature increased from 65 to 90°C. The activation energy for the initial deposition of electroless Co-P film was 15.67
kcal/mol, which was calculated according to the Arrhenius equation.
©
2004 The Electrochemical Society. ͓DOI: 10.1149/1.1793691͔ All rights reserved.
Manuscript submitted October 21, 2003; revised manuscript received April 12, 2004. Available electronically September 27, 2004.
Brenner and Riddell first discovered autocatalytic chemical
Experimental
deposition of cobalt and nickel films from aqueous solutions of the
metal chlorides or sulfates on a catalytic surface. Many other work-
Electroless cobalt depositions occur only on certain metallic sur-
faces such as iron, nickel, cobalt, and palladium. In this study, Si
wafer was used as a substrate. Because silicon wafers are semicon-
ductors, pretreatment is necessary. The substrate ͑Si wafer͒ was pre-
1
ers followed and developed the practice of electroless Co and Ni
plating for numerous applications such as magnetic recording me-
dia, metallization of plastics, and wear, corrosion-resistant
2
-5
sensitized by treatment with solution of SnCl /HCl (40 g/L SnCl2
surface.
2
ϩ 40 mL/L HCl͒ and preactivated by treatment with a solution of
Electroless cobalt-phosphorus ͑Co-P͒ films are interesting prima-
rily due to their magnetic properties. Therefore, the relationships
among the magnetic properties, the plating variables, and the micro-
structures of the electroless Co-P film have been studied
PdCl2 /HCl (0.15 g/L PdCl2 ϩ 3 mL/L HCl ϩ 83 mL/L HF) before
electroless plating. Prior to this. Si wafers were cleaned by ultra-
sonic vibration in acetone solution first, in solution of H SO /H O
2
4
2
2
4
,6-12
extensively.
There is little literature about the growth kinetics of
and then etched in 10% HF aqueous solution to remove oxide. Elec-
troless cobalt was deposited on Si wafers at various temperatures
and times. The plating temperature was 65, 75, 85, and 90°C. The
electroless solution decomposed at 95°C in this study. To find the
change of the morphology and thickness of the film, we imple-
mented a plating test at various plating temperatures and with vari-
ous plating times. The composition of the electroless cobalt plating
bath and plating parameters are listed in Table I. The ratio between
the electroless Co-P plating. Frieze et al. studied the nucleation of
1
3
electroless Co-P deposits on carbon film by electron microscopy.
They showed that Co-P particles nucleated on palladium-activated
surfaces at the beginning of deposition, and as the deposition pro-
gressed, the distribution of the Co-P particles became more
uniform.13 The size and density of Co-P particles depends on the
temperature of the bath with lower hypophosphite contents. Similar
results have been reported by Khan and Lee and Hwang and
3
2
Lin.14,15
the volume ͑cm ͒ of the plating bath and the total surface area ͑cm ͒
of substrate was 15.
They showed that electroless Co-P deposits on Al-based
Ni-P substrate initially nucleate in a three-dimensional islands form
at dispersive locations, coalesce together very slowly, and finally
become channel-like. Film deposited at 83°C and pH 8.05 becomes
continuous at a thickness of 30 nm and has a rough surface.
The microtopology, chemical composition, and crystal structure
of a specimen can be examined simultaneously by transmission
electron microscopy ͑TEM͒. TEM can offer the following merit for
the study of the growth kinetics of electroless Co-P deposition; that
is, the initial state of nucleation and growth before the deposits
become a continuous film can be clearly investigated. In this study,
we find the thickness of the deposited films by TEM. TEM can
measure thin films thickness down to several nanometers. The thick-
ness of electroless deposits is usually found by atomic
A JEOL-2010EX transmission electron microscope operating at
200 kV was used for TEM examination. The ͓110͔-oriented cross-
sectional transmission electron microscopy ͑XTEM͒ samples were
prepared following the procedures outlined by Sheng and Chang.
The chemical composition of the deposited materials was deter-
mined by energy dispersive X-ray spectroscopy ͑EDS͒ equipped
with a microscope.
Table I. Composition of electroless plating cobalt bath and plat-
ing parameters.
1
4,16
6,13,17
absorption,
resistance change,
and weight change
techniques.18 Among these methods for measuring the thickness of
the films, TEM examination is the most precise. In this study, the
thin Co-P film is deposited on a Si wafer only for a few minutes so
that the concentrations of electroless cobalt bath almost keep con-
stant. And the effect of plating temperature on the electroless cobalt
reaction can be studied without concerning the variance of bath
concentrations during electroless plating.
Composition of bath
and plating parameters
Specifications
CoSO •7H O ͑g/L͒
35
40
35
4
2
NaH PO ͑g/L͒
2
2
Na citrate ͑g/L͒
3
(
NH ) SO ͑g/L͒
70
4
2
4
This work is related to the investigation of electroless cobalt
deposition on silicon wafers in alkaline baths containing hypophos-
phite reduction agents by TEM. The activation energy and initial
growth of electroless cobalt thin films were studied.
PEG 2000 ͑ppm͒
ϳ20
pH
9.00 Ϯ 0.01
Plating temperature and times
90.0 Ϯ 0.5°C ͑1, 1.5, 2, 2.5 min͒
8
5.0 Ϯ 0.5°C ͑2, 3, 4, 5 min͒
5.0 Ϯ 1°C ͑4, 6, 8, 10 min͒
65.0 Ϯ 1°C ͑8, 11, 14, 17 min͒
7
z
E-mail: dafney@sunws.nhust.edu.tw

Journal of The Electrochemical Society, 151 ͑10͒ C680-C683 ͑2004͒
C681
Figure 2. BF images of XTEM micrographs of the as-deposited sample at
5°C for ͑a͒ 60 and ͑b͒ 120 s.
8
Results and Discussion
Figure 1a is the TEM plan view of the sample, the silicon with
100͒ surface was cleaned in HF solution, sensitized for 2 min in
SnCl solution, and activated for 30 s in PdCl solution. The single
͑
2
2
Pd particle is several nanometers large and the Pd particles tend to
aggregate in clusters. Generally, the clusters dispersed on the sample
surface evenly. The TEM diffraction pattern of the sample in Fig. 1a
is shown in Fig. 1b. The regular spots pattern belongs to single-
crystal silicon and the ring pattern is formed by tiny Pd particles.
Figure 2a shows the XTEM micrograph of a Co-P deposit plated at
85°C for 60 s. The Co-P deposit is composed of different size par-
ticles. It is clear that the nucleation mechanism of electroless Co-P
films corresponds to an island growth model. The result is the same
as electroless Co-P deposits on Al-based Ni-P substrate that initially
nucleate in a three-dimensional ͑3D͒ islands form at dispersive lo-
cations. The Co-P deposit cannot become a continuous film until
plating time is long enough. The Co-P deposit at a plating tempera-
ture 85°C for 120 s becomes a continuous film 173 nm thick and has
an uneven surface, as shown in Fig. 2b. To evaluate the effect of
plating temperature on the Co-P deposit, the morphology of the
deposit plated at 65, 75, and 90°C for 60 s was also examined. The
deposit plated at 90°C for 60 s was continuous. But the deposit
plated at 85°C for 60 s was discontinuous. And there was no Co-P
deposit on a Si wafer plated at 75 and 65°C for 60 s. Co-P deposits
were found on Si wafers at plating temperatures of 75 and 65°C with
a plating time longer than 60 s. The nucleation of a Co-P deposit is
inhibited by decreasing the plating temperature. The results reveal
that the Co-P deposits initially nucleate into particles with different
sizes and shapes, coalesce together sequentially, and finally become
a continuous film. Clearly, the least plating time for the Co-P deposit
to become a continuous film decreases with plating temperature in-
creasing from 65 to 90°C at the same bath pH, as shown in Table II.
Co-P films deposited at different temperatures have similar compo-
sitions with about 8.0 atom % P content that is detected by TEM/
EDS.
Figure 1. Bright-field ͑BF͒ images of TEM micrographs of the silicon wafer
after sensitized and activated treatment; ͑a͒ plan view and ͑b͒ diffraction
pattern.
Figures 3a-d show the XTEM micrographs of Co-P films plated
Table II. Average thickness of electroless cobalt films.
90°C
85°C
75°C
65°C
Thickness of
electroless
cobalt films
͑nm͒
Thickness of
electroless
cobalt films
͑nm͒
Thickness of
electroless
cobalt films
͑nm͒
Thickness of
electroless
cobalt films
͑nm͒
Plating
time ͑s͒
Plating
time ͑s͒
Plating
time ͑s͒
Plating
time ͑s͒
6
9
20
50
0
0
155.8
245.8
327.6
397.3
120
180
240
300
173.3
297.8
430.3
578.7
240
360
480
600
328.1
468.2
631.3
751.3
480
660
840
240.2
321.9
420.0
526.0
1
1
1020

C682
Journal of The Electrochemical Society, 151 ͑10͒ C680-C683 ͑2004͒
Figure 3. BF images of XTEM micrographs of the as-deposited sample at
a͒ 65°C, 8 min, ͑b͒ 75°C, 8 min, ͑c͒ 85°C, 2 min, and ͑d͒ 90°C, 2 min.
͑
at 65, 75, 85, and 90°C, respectively. The plating time was 8 min for
the Co-P films in Fig. 3a and b and 2 min for the Co-P films in Fig.
3c and d. The Co-P deposits are all crystalline and are composed of
grains with different sizes and shapes. The crystalline phases are
confirmed by electron diffraction. The thickness of Co-P films in-
creases as plating temperature increases for the same plating time
and the surfaces of Co-P films are uneven. The XTEM micrographs
in Fig. 4a-u ͑a-u means upper portion image in a͒, b-u, c-u, and d-u
show the growing morphology of Co-P films plated at 75°C for 4, 6,
8, and 10 min, respectively. To measure the thickness of Co-P de-
posits with uneven surfaces, we developed a method consisting of
the following steps: ͑i͒ reproducing the contour of the Co-P deposit
on graph paper as shown in Fig. 4a-l ͑a-l means lower portion image
in a͒, b-l, c-l, and d-l; (ii) measuring the area inside the contour of
the Co-P deposits; and (iii) getting the thickness of the Co-P de-
posits by dividing the area by the width of the contour.
The average thickness of electroless cobalt films for various tem-
peratures and times are shown in Table II. The average thickness of
the electroless cobalt films is in the range of 155-728 nm. The rela-
tion of the average film thickness vs. plating time is shown in Fig. 5.
The relation curves are almost linear and the coefficient of determi-
nation of each fitting line is over 0.995. The plating rates at different
temperatures can be determined from the slopes in Fig. 5. The plat-
ing rates are 0.54, 1.3, 2.2, and 2.7 nm/s for Co-P films plated at 65,
Figure 4. Contour of BF images and XTEM micrographs of the as-
deposited sample at 75°C, for ͑a͒ 4, ͑b͒ 6, ͑c͒ 8, and ͑d͒ 10 min.
Electroless cobalt plating is a catalytic reaction. It requires en-
ergy ͑temperature͒ to proceed. According to an Arrhenius equation,
a plot of ln ͑plating rate͒ vs. 1/T was used to determine the activa-
tion energy. A plot of ln ͑plating rate͒ vs. the reciprocal of the ab-
75, 85, and 90°C, respectively. The result reveals that the plating
rates increase with plating temperature increasing from 65 to 90°C.

Journal of The Electrochemical Society, 151 ͑10͒ C680-C683 ͑2004͒
C683
Figure 6. Plot of ln ͑rate͒ vs. 1/T.
References
Figure 5. Plot of film thickness vs. plating time.
1
2
. A. Brenner and G. E. Riddell, RP1725, J. Res. Natl. Bur. Stand., 37, 31 ͑1946͒.
. N. Petrov, Y. Sverdlov, and Y. Shacham-Diamand, J. Electrochem. Soc., 149, C187
͑2002͒.
solute temperature, 1/T, is shown in Fig. 6 and the coefficient of
determination of the fitting line is 0.98. The activation energy for the
initial deposition of electroless Co-P film is 15.67 kcal/mole. G. O.
Mallory and J. B. Hajdu reported the activation energy of electroless
nickel reaction with citrate complexing agents is 16 kcal/mole by
using weight change to determine the plating rate ͑␮m/h͒. The
result reveals that both electroless reactions have nearly equal acti-
vation energy for growth.
3. T. Osaka, Electrochim. Acta, 42, 3015 ͑1997͒.
4
5
. T. Chen, D. A. Rogowski, and R. M. White, J. Appl. Phys., 49, 1816 ͑1978͒.
. N. M. Martyak, S. Wetterer, L. Harrison, M. Mcneil, R. Heu, and A. Albert
Neiderer, Plat. Surf. Finish., 81, 60 ͑June 1993͒.
1
9
6. D. E. Speliotis, J. R. Morrision, and J. S. Judge, IEEE Trans. Magn., MaG-1, 348
͑1965͒.
7
8
. R. O. Cortijo and M. Schlesinger, J. Electrochem. Soc., 131, 2800 ͑1984͒.
. M. Mirzamaani, L. Romankiw, C. Mcgrath, and J. Karasinski, J. Electrochem. Soc.,
1
35, 2813 ͑1988͒.
Conclusions
9. D. Dimilia, J. Horkans, C. McGrath, M. Mirzamaani, and G. Scilla, J. Electrochem.
Soc., 135, 2817 ͑1988͒.
We conclude from the results in this study: ͑i͒ electroless cobalt
deposits on silicon with an alkaline bath containing hypophosphite
reduction agents initially nucleate into particles with different sizes
and shapes, coalesce together sequentially, and finally become a
continuous film; (ii) the surface of the electroless cobalt films is
uneven; (iii) the plating rate of the Co-P films is 0.54 nm/s ϳ2.7
nm/s and linearly increases as temperature increases from 65 to
1
0. K. Hono and D. E. Laughlin, J. Magn. Magn. Mater., 80, L137 ͑1989͒.
11. M. Schlesinger, X. Meng, W. T. Evans, D. A. Saunders, and W. P. Kampert, J.
Electrochem. Soc., 137, 1706 ͑1990͒.
1
2. R. Tarozaite, M. Kurinatiene, A. Dziuve, and Z. Jusys, Surf. Coat. Technol., 115, 57
1999͒.
3. A. S. Frieze, R. Sard, and R. Weil, J. Electrochem. Soc., 115, 586 ͑1968͒.
͑
1
14. M. R. Khan and J. I. Lee, J. Appl. Phys., 57, 15, 4028 ͑1985͒.
1
1
1
5. B. J. Hwang and S. H. Lin, J. Electrochem. Soc., 142, 3749 ͑1995͒.
6. E. L. Nicholoson and M. R. Khan, J. Electrochem. Soc., 133, 2342 ͑1986͒.
7. F. M. Donahue and C. U. Yu, Electrochim. Acta, 15, 237 ͑1970͒.
9
0°C; and (iv) the activation energy for initial deposition of elec-
troless Co-P films on silicon wafers is 15.67 kcal/mol.
18. A. Malecki and A. Micek-Ilnicka, Surf. Coat. Technol., 123, 72 ͑2000͒.
1
9. G. O. Mallory and J. B. Hajdu, Electroless Plating: Fundamentals and Application,
p. 98, The American Electroplaters and Surface Finishers Society, Orlando, FL
͑1990͒.
National Huwei University of Science and Technology assisted in meet-
ing the publication costs of this article.