Effects of polyethylene glycol and gelatin on the crystal size, morphology, and Sn2+-sensing ability of bismuth deposits

Article abstract of DOI:10.1016/j.electacta.2011.06.077

The influences of citric acid (CA), ethylenediaminetetraacetic acid (EDTA), polyethylene glycol (PEG), and gelatin on the deposition behavior of Bi were systematically investigated through the linear sweep voltammetric (LSV) analysis. Based on the LSV res

Full text of DOI:10.1016/j.electacta.2011.06.077

Electrochimica Acta 56 (2011) 7615–7621
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Effects of polyethylene glycol and gelatin on the crystal size, morphology, and
2
+
Sn -sensing ability of bismuth deposits
2
2
∗,1
Yi-Da Tsai , Chein-Hung Lien , Chi-Chang Hu
Department of Chemical Engineering, National Tsing Hua University, Hsin-Chu 30013, Taiwan
a r t i c l e i n f o
a b s t r a c t
Article history:
The influences of citric acid (CA), ethylenediaminetetraacetic acid (EDTA), polyethylene glycol (PEG),
and gelatin on the deposition behavior of Bi were systematically investigated through the linear sweep
voltammetric (LSV) analysis. Based on the LSV results, deposits plated from a typical solution contain-
ing 0.05 M Bi(NO3)3·5H2O and various combinations of complex agents and additives with pH = 3.5 at 1
Received 28 January 2011
Received in revised form 27 May 2011
Accepted 25 June 2011
Available online 1 July 2011
−2
and 30 mA cm were characterized by scanning electron microscopic (SEM) and X-ray diffraction (XRD)
analyses. The adhesion of deposits and the formation of dendrites were respectively improved and inhib-
ited by the adsorption of PEG onto Bi deposits. With adding the above four compounds, a synergistic
Keywords:
Bi sensor
Electrodeposition
Complex agent
Linear sweep voltammetry
Square-wave anodic stripping voltammetry
−
2
effect was shown to reach a nano-sized, sphere-like, porous morphology of a Bi deposit at 30 mA cm
.
The crystal size and morphology of Bi deposits were found to affect the sensing ability of Sn2+ through
the square-wave anodic stripping voltammetric (SWASV) analysis.
© 2011 Elsevier Ltd. All rights reserved.
1
. Introduction
Recently, increased attention on the electrochemical deposition
of the mercury-free electrodes with environmental friendly proper-
ties, which are comparable to MFEs, have been proposed to replace
MFEs for the application of sensors.
of bismuth has been paid for the electrochemical community, due
to the unique electrical and physicochemical properties of bismuth
for several applications [1–6], especially in developing the heavy
metal sensors [7,8]. For instance, an extremely high magneto-
resistance effect has been observed earlier in the Bi single-crystals,
thin films, and nanowires, leading Bi to be a promising material
of the magnetic sensor. In addition, bismuth is an environment
friendly element with a low toxicity, which has been used widely
in the pharmaceutical industry [1]. Besides, bismuth film elec-
trodes (BFEs) were proposed for the cathodic detection of organic
compounds and metallic ions [2], the anticorrosive interlayer in
batteries [3], magnetic field [4], current sensors [5], and decorative
coatings due to the excellent hydrogen embitterment inhibition of
Bi coatings [6]. Traditionally, determination of metals or heavy met-
als was extensively carried out at mercury-film electrodes (MFEs)
by means of the anodic stripping voltammetric analysis, due to the
advantages of high sensitivity, reproducibility, and high hydrogen
evolution overpotential, etc. [7,8]. However, because of its toxicity
to human health and the damages to our environment, mercury-
based electrodes must be restricted. Consequently, BFEs, one kind
In general, BFEs can be fabricated by several methods, such as
thermal evaporation [9], electrodeposition [10,11], DC sputtering
[12], RF magnetron sputtering [13], and molecular beam epitaxy
[14]. Among these methods, electrodeposition is a simple route
to control the surface morphologies of the coatings via changing
the electroplating solutions such as composition, complex agents,
additives, pH, and temperature, as well as the deposition parame-
ters e.g., current density and applied potential [15–17]. Moreover,
it offers other merits, including a high deposition rate, mass pro-
ductivity, industrially well-established technology, cost-effectivity,
etc. Hence, electrodeposition has been progressively recognized as
a popular technology for the coating application. On the other hand,
little literature has been published on how to control the morphol-
ogy and adhesion of Bi deposits through electroplating [10,11],
probably due to the instability of free Bi3+ ions. Accordingly, one
of the purposes in this study is to control the morphology and
crystal size of Bi deposits through electroplating under a DC mode
by adding organic agents in various combinations into the plating
solution.
In this work, the influences of citric acid (CA), ethylene-
diaminetetraacetic acid (EDTA), polyethylene glycol (PEG), and
gelatin on the deposition behavior of Bi were systematically investi-
gated through the linear sweep voltammetric (LSV) analysis. Based
on the LSV results, Bi deposits were electroplated from the bath
containing different combinations of complex agents and additives.
The morphologies and crystalline structure of resultant deposits
∗ Corresponding author. Tel.: +886 3 5736027; fax: +886 3 5736027.
E-mail address: cchu@che.nthu.edu.tw (C.-C. Hu).
ISE Member.
Both the authors contributed equally to this work.
1
2
0
013-4686/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2011.06.077

7
616
Y.-D. Tsai et al. / Electrochimica Acta 56 (2011) 7615–7621
Table 1
The bath composition and plating conditions of Bi deposition.
Bi(NO3)3·5H2O
0.05
0.3
M
M
Citric acid (CA)
Ethylenediaminetetraacetic acid (EDTA)
0.3
M
Polyethylene glycols (PEG, MW 400)
0.2
M
Gelatin
4000
1 or 30
25
ppm
mA cm
C
−
2
Current density
Temperature
pH
◦
3.5
were characterized by means of scanning electron microscopic
SEM) and X-ray diffraction (XRD) analyses, respectively. Finally,
(
the sensitivity of five typical BFEs on Sn(II) is evaluated by using
the square-wave anodic stripping voltammetry (SWASV) method.
In fact, the LSV responses of Bi deposition from the plating solutions
containing CA, EDTA, and PEG have been preliminarily investigated
in our previous studies for Sn–Bi alloy deposition [15,17]. However,
the effects of these organic agents in addition to gelatin on the Bi
deposition behavior in previous baths are not directly applicable to
this study because of the significant differences in pH values among
these plating baths.
Fig. 1. LSV curves measured from the plating solution containing 0.05 M Bi3+ with
1) 0.3 M CA, (2) 0.3 M CA + 0.2 M PEG, (3) 0.3 M CA + 4000 ppm gelatin, and (4)
.3 M CA + 0.2 M PEG + 4000 ppm gelatin onto the Cu/Ni electrode. LSV curves were
(
0
−1
◦
measured at 8 mV s and 25 C.
tion (pH = 2) without stirring. The SWASV curves were recorded
between −0.6 and −0.3 V at an applied frequency of 15 Hz with the
amplitude of 25 mV and a potential step of 4 mV.
2
. Experimental
Bismuth deposits were electroplated from a simple solution
Morphologies of all deposits were examined by a field emis-
sion scanning electron microscope (FE-SEM, Hitachi S-4700). X-ray
diffraction patterns were obtained from an X-ray diffractometer
(CuK␣, Ultima IV, Rigaku).
−2
2
−1
onto copper (99.5%, 1.094 × 10 cm , 1.18 × 10 cm in diam-
eter) circular plates which have been deposited with a nickel
film (ca. 2 ␮m). The pretreatment procedures of Cu/Ni substrates
completely followed our previous work [15–19]. The Cu/Ni sub-
strates, rinsed with deionized water, were vertically placed in a
All solutions were prepared with deionized water produced by
a reagent water system (Milli-Q SP, Japan) at 18 Mꢀ cm and all
reagents not specified were Merck, GR. The solutions for the LSV
and SWASV analyses were degassed with purified nitrogen gas
for 20 min before the electrochemical measurements and nitrogen
was passed over the solution during the measurements. Solution
temperature was maintained at the specified temperature with an
1
(
00-ml jacket cell surrounded with a dimensionally stable anode
®
DSA , an (Ru–Ti)O -coated Ti electrode). All substrates were
2
electroplated with Bi deposits with a constant passed charge den-
−2
2+
sity (0.09 C cm ) for the textural analyses and the Sn -sensing
tests. The basic plating bath, shown in Table 1, consisted of
◦
0
.05 M Bi(NO ) ·5H O (Hayashi, EP) and various complex agents
accuracy of 0.1 C by means of a water thermostat (Haake DC3 and
3
3
2
or additives, including citric acid (CA, Shimakyu, EP), ethylene-
diaminetetraacetic acid (EDTA, Riedel-deHaen, GA), polyethylene
glycols (PEG 400, MW 400, Shimakyu, EP), and gelatin (from porcine
skin, Type A∼300 Bloom (G2500-100G)) in various combinations.
The pH of plating baths was adjusted to be 3.5 with concentrated
K20).
3. Results and discussion
3.1. Linear sweep voltammetric study
HCl or NH OH. The plating baths were not agitated during the DC
4
electroplating of Bi. After deposition, these electrodes were repeat-
edly rinsed with deionized water and finally dried in a vacuum
oven at room temperature. The thickness of Bi deposits, signifi-
cantly influenced by the combination of organic agents employed
in this work, is between 3.9 and 0.5 ␮m although a constant passed
Curves 1–4 in Fig. 1 show the influences of PEG and gelatin
on the deposition behavior of Bi on the Cu/Ni substrate in the
solution containing 0.05 M Bi3+ and 0.3 M CA. On curve 1, Bi depo-
sition commences at ca. −240 mV with an obvious peak centered
3+
at ca. −280 mV, attributable to the deposition of free Bi with a
−2
−2
charge density (0.09 C cm ) is employed.
limiting current density of 12 mA cm from −400 and −600 mV.
The electrochemical analyses were performed with an electro-
chemical analyzer system, CHI 660c (CH Instruments, USA). All
electrochemical analyses were carried out in a three-compartment
cell. An Ag/AgCl electrode (Argenthal, 3 M KCl, 0.207 V versus a
The shoulder wave at ca. −1000 mV may correspond to the Bi
deposition from the Bi3+–CA complex species, indicating a weak
complex agent of citrate for Bi3+ [15,17]. The increase in the cur-
rent density at potentials negative to −600 mV is attributed to the
hydrogen evolution reaction (HER). On curve 2, the limiting cur-
◦
standard hydrogen electrode (SHE) at 25 C) was utilized as the
−
2
3+
reference electrode and a piece of platinum gauze with an exposed
area equal to 4 cm2 served as the counter electrode. A Luggin cap-
illary, whose tip was set at a distance of 1–2 mm from the surface
of substrates, was used to minimize errors due to iR drop in the
electrolytes. For the SWASV measurements, the testing solutions
contain 0.1 M CA (Shimakyu, EP) and SnCl ·nH O (Hayashi, EP) in
rent density (ca. 10 mA cm ) of free Bi deposition continues to
−1200 mV, suggesting a leveling effect of this strongly adsorbed
PEG, leading to a uniform deposit [17,20]. The above different i–E
responses at potentials negative to −400 mV are attributed to the
PEG adsorption on Bi already deposited, inhibiting the deposition
3+
of Bi –CA complex species. On the other hand, the sharp increase
in the hydrogen evolving currents behind −1400 mV implies the
potential-dependent desorption of PEG on Bi and Cu/Ni substrate
in this potential region [21]. On curve 3, gelatin is considered as a
leveling agent due to its strong adsorption ability [22] and a com-
plex agent coordinated with metallic ions [22,23]. The similar i–E
responses of curves 1 and 3 at potentials positive to −600 mV sug-
2
2
various concentrations. The concentrations of dissolved Sn(II) ions
have been confirmed by an inductively coupled plasma-mass spec-
trometer (ICP-MS, SCIEX ELAN 5000). The cathodic deposition for
accumulating Sn onto the BFEs was fixed at −0.8 V for 300 s under
a magnetic stirring rate of 100 rpm. The Sn-coated BFEs rinsed
with deionized water were anodically stripped in a 0.1 M CA solu-

Y.-D. Tsai et al. / Electrochimica Acta 56 (2011) 7615–7621
7617
this organic combination on the Bi deposition and HER. Two unobvi-
ous peaks around −1170 and −1230 mV may be due to the different
deposition potentials of Bi3+–EDTA complexes in various structures
when gelatin leaves the substrate at ca. −1150 mV. Since Bi depo-
sition and HER occur in the same potential region, gelatin should
mainly act as an adsorption additive in plating Bi due to the strong
coordination between Bi3+ and EDTA. A similar phenomenon is also
visible for adding PEG into this bath (curve 2) while the shield-
ing effect of gelatin on Bi deposition and HER is much stronger
and more complete than that of PEG, judged from the more nega-
tive onset potential of Bi deposition and the double-layer-charging
response at potentials above −1150 mV on curve 3. Thus, a stronger
synergistic effect on suppressing the Bi deposition is obtained for
this organic combination. On curve 4, Bi deposition commences at
−
2
ca. −700 mV (based on 0.5 mA cm ), a potential very close to that
−
2
on curve 2, but a much smaller plateau (1.7 mA cm ) is obtained
at about −850 mV. This plateau may correspond to the same reac-
tion in the peak centered at −1000 mV on curve 2. Therefore, the
unchanged deposition potential of Bi3+–EDTA but a lower current
density indicates a significant suppression in the Bi deposition rate
through adding gelatin into the CA + EDTA + PEG solution.
Fig. 2. LSV curves measured from the plating solution containing 0.05 M Bi3+ and
.3 M EDTA with (1) 0.3 M CA, (2) 0.3 M CA + 0.2 M PEG, (3) 0.3 M CA + 4000 ppm
gelatin, and (4) 0.3 M CA + 0.2 M PEG + 4000 ppm gelatin onto the Cu/Ni electrode.
0
−1
◦
LSV curves were measured at 8 mV s and 25 C.
gest that gelatin is a weak complex agent for Bi3+. The relatively
3.2. Textural analyses
−
2
higher limiting current density, ca. 15 mA cm between −600 and
−
1200 mV, suggests weaker adsorption strength of gelatin in com-
paring with PEG.
Based on the LSV analyses, the crystal size and morphology of
Bi deposits can be controlled by changing the combination of com-
plex agents and additives. Figs. 3 and 4 show the SEM morphologies
of Bi deposits (at 5000 and 200,000 magnifications, respectively)
plated from the baths containing various combinations of complex
agents/additives at two current densities. The adhesion, testing by
ultrasonic vibration, of these deposits on the Cu/Ni substrate is
shown in Table 2. From an examination of all SEM images, several
distinct morphologies were obtained, which are strongly depen-
dent on the composition of plating baths and the current density of
deposition. In Figs. 3a, e and 4a, e, both Bi deposits show dendrite-
Curve 4 in Fig. 1 demonstrates the synergistic effect of the simul-
taneous addition of 0.3 M CA, 0.2 M PEG, and 4000 ppm gelatin
into the electroplating bath. Here, gelatin, a polypeptide with high
molecular mass, may form microspheres, adheres together, and dis-
plays a poor flow property in the gelatin–water system. Hence,
gelatin usually acts as a film-forming polymer/surfactant to con-
trol the surface tension, which promotes the emulsification in the
photographic industry [24]. However, the addition of PEG into the
gelatin–water solution avoids this gathering phenomenon among
the gelatin microdrops in the internal phase of emulsion by electro-
static and hydrophobic forces [25]. Thus, PEG works as a physical
barrier for the aggregation of gelatin microspheres. As a result,
some new micro-sized PEG–gelatin species should co-work to form
a thin layer adsorbed onto the substrate/deposit surface [24,26].
Hence, curve 4 shows a negative shift in the onset potential to
−
2
like morphologies when the current density is 30 mA cm . The
adhesion of both deposits is very poor (see Table 2), indicating
−
2
that the deposition rate at 30 mA cm is too high. When a current
−
2
density of 1 mA cm is applied, the deposit plated from the bath
containing both CA and EDTA shows a relatively compact morphol-
ogy with good adhesion. Under such a deposition current density,
the stacking of Bi atoms belongs to the pyramidal growth, indicat-
ing a steady growth of Bi crystals [27,28]. However, the Bi deposit
plated from the solution containing CA only shows the dendrite
morphology with poor adhesion although the current density is
−
2
ca. −380 mV, a lower deposition rate (peak current ≈ 17 mA cm ),
and the lowest limiting current density among all baths. This
result implies the formation of a relatively stable metallic complex
3
+
species, Bi –PEG–gelatin.
−
2
Curves 1–4 in Fig. 2 show the effects of PEG and gelatin on the
Bi deposition behavior measured from a bath containing 0.05 M
1 mA cm . Due to the rapid nucleation of Bi clusters, dendrites are
easily obtained for a Bi deposit plated from a simple bath contain-
3+
3+
3+
Bi + 0.3 M CA + 0.3 M EDTA. A comparison of curve 1 in Figs. 1 and 2
reveals that EDTA in the electroplating bath not only causes a neg-
ative shift in the onset potential of Bi deposition from −240 to
ing Bi only [10,15,17,29]. Thus, CA, a weak complex agent for Bi ,
3+
cannot form stable Bi –CA species in the plating bath at pH = 3.5
to reduce the nucleation rate of Bi [15]. The above difference sug-
gests that the strong coordination between EDTA and Bi3+ not only
negatively shifts the onset deposition potential but also reduce the
nucleation rate of Bi clusters.
−
550 mV but also clearly reduces the peak current density of Bi
deposition, indicative of relatively strong coordination between
3+
3+
Bi and EDTA. This result, similar to that measured from a Bi
solution at pH = 2 or 6, indicates that combination of CA and EDTA
can significantly coordinate with Bi3 [15–17]. Note the presence
of a shoulder and a peak on curve 1, suggesting that the concen-
tration of EDTA is not sufficient to form stable Bi3 –EDTA complex
species. Thus, much more polarizable behavior of Bi deposition is
obtained on curve 2 in Fig. 2. Clearly, the significant negative shift
in the deposition potential and a much lower peak current density
can be achieved by the coexistence of CA, EDTA, and PEG, reveal-
ing a synergistic effect on suppressing the deposition rate of Bi (in
comparing with curve 2 in Fig. 1 and curve 1 in Fig. 2), similar to
curve 4 in Fig. 1.
From Fig. 3b, the Bi deposit plated from the solution containing
CA and PEG is piled with many grains without obvious dendrites.
Under a high magnification (see Fig. 4b), these grains are composed
of short rods and screw dislocations, indicating the enrichment of
imperfections while this deposit shows excellent adhesion on the
+
+
−
2
substrate. Besides, the deposits plated at 1 or 30 mA cm show a
similar morphology, indicating steady growth and nucleation of Bi
−
2
grains although the HER occurs significantly at 30 mA cm . From a
comparison of Fig. 3a and b (or Fig. 4a and b), the relatively compact,
polycrystalline, smooth morphology without dendrites reveals the
significant improvement in the physical properties of resultant
deposits with good adhesion, due to the strong adsorption of PEG on
the deposit [15]. In Figs. 3f and 4f, the Bi deposit shows a branch-like
On curve 3, no significant current is found at potentials positive
to ca. −1150 mV, indicating a nearly complete shielding effect of

7
618
Y.-D. Tsai et al. / Electrochimica Acta 56 (2011) 7615–7621
−
2
Fig. 3. SEM morphologies under a magnification of 5000 for Bi deposits plated at (1) 1 and (2) 30 mA cm from the basic plating bath (0.05 M Bi(NO3)3) with (a) 0.3 M CA,
b) 0.3 M CA + 0.2 M PEG, (c) 0.3 M CA + 4000 ppm gelatin, (d) 0.3 M CA + 0.2 M PEG + 4000 ppm gelatin, (e) 0.3 M CA + 0.3 M EDTA, (f) 0.3 M CA + 0.3 M EDTA + 0.2 M PEG, (g)
.3 M CA + 0.3 M EDTA + 4000 ppm gelatin, and (h) 0.3 M CA + 0.3 M EDTA + 0.2 M PEG + 4000 ppm gelatin.
(
0
morphology enriched with thin skeletons when the applied current
by adding PEG and that achieving a smoother film due to the PEG
adsorption becomes more obvious at a higher current density of
deposition.
In Fig. 4c and g, the Bi deposits plated from the solution with
the gelatin additives show smoother surface morphologies than
that plated from similar baths containing the PEG additives. This
−2
density is 1 mA cm . On the contrary, as the applied current den-
−2
sity is 30 mA cm , the relatively smooth morphology is replaced
with a relatively fine, star-like structure. The above improvement
in morphology and adhesion in comparison with Figs. 3e and 4e
strongly suggests that the dendrite formation can be suppressed
Table 2
Summary of electroplating conditions from the plating solution containing 0.05 M Bi(NO3)3 and resultant properties of Bi deposits.
Compounds
Current density (mA cm−2)
Adhesion
Avg. crystal sizea (nm)
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
a1) CA
a2) CA
b1) CA + PEG
b2) CA + PEG
c1) CA + gelatin
c2) CA + gelatin
d1) CA + PEG + gelatin
d2) CA + PEG + gelatin
e1) CA + EDTA
1
30
1
30
1
30
1
30
1
30
1
30
1
30
1
30
Poor
Poor
Excellent
Excellent
Poor
–
–
43.7
38.9
–
Poor
–
Excellent
Excellent
Very good
Poor
Very good
Very good
Poor
Poor
Very good
Very good
38.1
27.1
67.6
–
49.7
34.4
–
–
31.4
29.8
e2) CA + EDTA
f1) CA + EDTA+ PEG
f2) CA + EDTA + PEG
g1) CA + EDTA + gelatin
g2) CA + EDTA + gelatin
h1) CA + EDTA + PEG + gelatin
h2) CA + EDTA + PEG + gelatin
CA: citric acid, 0.3 M. EDTA: ethylenediaminetetraacetic acid, 0.05 M. PEG: polyethylene glycol (MW 400), 0.2 M. Gelatin: 4000 ppm.
a
The average crystal size is estimated from the XRD peak with the highest intensity.

Y.-D. Tsai et al. / Electrochimica Acta 56 (2011) 7615–7621
7619
−
2
Fig. 4. SEM morphologies under a magnification of 200,000 for Bi deposits plated at (1) 1 and (2) 30 mA cm from the basic plating bath (0.05 M Bi(NO3)3) with (a) 0.3 M
CA, (b) 0.3 M CA + 0.2 M PEG, (c) 0.3 M CA + 4000 ppm gelatin, (d) 0.3 M CA + 0.2 M PEG + 4000 ppm gelatin, (e) 0.3 M CA + 0.3 M EDTA, (f) 0.3 M CA + 0.3 M EDTA + 0.2 M PEG, (g)
0
.3 M CA + 0.3 M EDTA + 4000 ppm gelatin, and (h) 0.3 M CA + 0.3 M EDTA + 0.2 M PEG + 4000 ppm gelatin.
result also indicates a strong adsorption effect of gelatin on the
surface of Bi deposit. However, both Bi films show a poor adhesion
property, full of cracks (see Fig. 3c and g), probably due to the for-
mation of gelatin microspheres which tend to adhere together and
can be considered as a compact shielding layer [24]. Since inhibi-
tion of the HER by the gelatin adsorption is weaker in comparison
with PEG (see Figs. 1 and 2), hydrogen bubbles should extensively
evolve at the boundaries of these compact shielding microspheres
during electroplating, resulting in the formation of deep cracks. In
Figs. 3d and 4d, the Bi deposits plated from the solution contain-
ing CA, PEG, and gelatin show very smooth surface morphologies
and relatively smaller grains in different microstructures, such as
flower-like or stitch-like, are homogeneously distributed on the
deposits by varying the plating current density. This result strongly
displays a synergistic adsorption effect of the mixed additives (i.e.,
a thin PEG–gelatin layer) on the deposits. In addition, in compari-
son with the deposits shown in Fig. 3c, the better adhesion of these
deposits is attributed to the physical interaction between PEG and
gelatin, decreasing the tendency to form a compact shielding layer
of gelatin microspheres. This phenomenon should avoid the much
localized hydrogen evolution on the deposit and consequently, the
binary additives effectively avoid the formation of cracks. A simi-
lar synergistic effect is also visible in Figs. 3h and 4h which show
the stitch-like and sphere-like morphologies with very good adhe-
sion when the current densities of electroplating are equal to 1 and
cussion, the deposits plated from the baths containing unitary and
binary additives (i.e., PEG, gelatin, or PEG–gelatin) show smoother
morphologies in comparison with their counterparts especially
−
2
under the plating current density of 30 mA cm
.
Fig. 5 shows the XRD patterns of the deposits with the crystalline
structure, indicated in Table 2. From Table 2 and Fig. 5, the deposits
with good adhesion generally show obvious diffraction peaks on
their XRD patterns. An examination of these XRD patterns reveals
that the preferred orientation of Bi deposits is significantly affected
by varying the combination of complex and additives. For exam-
ple, the (0 1 2) face is the preferred orientation for the deposits
plated from the bath containing CA and PEG, which is indepen-
dent of the plating current density (see curves 1 and 2). When
gelatin was added into the above bath, this crystal face is signif-
icantly depressed when the current density was increased from 1
−
2
to 30 mA cm and the (1 1 0) face becomes the preferred orien-
tation (see curves 3 and 4). For the deposits plated from the bath
containing CA, EDTA, and PEG (see curves 5 and 6), the diffrac-
tion intensities of all peaks are low although the intensity of the
(0 1 2) face is the highest. A similar phenomenon is also visible
for the deposits plated from the bath containing CA, EDTA, PEG,
and gelatin (see curves 7 and 8) while the preferred orientation is
slightly changed by varying the current density of deposition. On
−
2
curve 9, the deposit plated at 1 mA cm from the bath contain-
ing both CA and EDTA shows a polycrystalline structure with an
excellent crystalline quality. This result supports the statement that
−2
3
0 mA cm , respectively. Based on all the above results and dis-

7
620
Y.-D. Tsai et al. / Electrochimica Acta 56 (2011) 7615–7621
Fig. 5. XRD patterns of Bi deposits with good adhesion property plated from the basic plating bath with (1) 0.3 M CA + 0.2 M PEG, at 1 mA cm−2, (2) 0.3 M CA + 0.2 M PEG, at
−
2
−
2
−
2
3
1
0 mA cm , (3) 0.3 M CA + 0.2 M PEG + 4000 ppm gelatin, at 1 mA cm ; (4) 0.3 M CA + 0.2 M PEG + 4000 ppm gelatin, at 30 mA cm ; (5) 0.3 M CA + 0.3 M EDTA + 0.2 M PEG, at
−
2
−2
−2
mA cm , (6) 0.3 M CA + 0.3 M EDTA + 0.2 M PEG, at 30 mA cm ; (7) 0.3 M CA + 0.3 M EDTA + 0.2 M PEG + 4000 ppm gelatin, at 1 mA cm , (8) 0.3 M CA + 0.3 M EDTA + 0.2 M
PEG + 4000 ppm gelatin, at 30 mA cm 2, and (9) 0.3 M CA + 0.3 M EDTA, at 1 mA cm−2
−
.
under this low deposition current density, the stacking of Bi atoms
belongs to the pyramidal growth and a steady growth of Bi crystals
[
27,28]. Based on the full width at half-maximum (FWHM) method,
the average crystal size of the Bi deposits was estimated from the
strongest peak on all XRD patterns. These results are shown in
Table 2. Obviously, the crystal size of Bi deposits plated from the
same bath is larger when the current density of deposition is lower,
reasonably due to a longer time and more opportunities for the
surface movement of Bi ad-atoms freshly deposited on the surface
of Bi grains. Accordingly, the crystalline quality and size of Bi are
improved.
3
.3. Sn2+ sensing by means of the SWASV analysis
This section tries to demonstrate the influences of certain
physicochemical properties such as morphology and crystal size of
Bi deposits (i.e., BFEs) on their sensing ability of Sn2 ions. Here, five
Bi deposits with their crystal size ∼68, ∼39, ∼34, ∼30, and ∼27 nm
were tested for this purpose. Since the background responses of
BFEs in the stripping solution are very important to the sensing
performances, the SWASV curves for the above five BFEs as well
as the Ni/Cu substrate and a BFE pre-deposited with Sn are shown
in Fig. 6a. From curve 6 in Fig. 6a, Ni starts to be oxidized at about
+
−
200 mV and a peak centered at 20 mV is found. For BFEs, their dis-
tinct stripping potential window can be set from ca. −300 to 0 mV
with a peak centered at −80 mV to avoid the possible interference
from Ni stripping although the difference between curves 1 and
6
is relatively unobvious. On the other hand, this effect shows no
2
+
influence on the Sn sensing since on curve 7, the potential win-
dow for anodic stripping of Sn is from ca. −600 to −300 mV with
a peak at −480 mV. From all the above results and discussion, the
SWASV curves for the anodic stripping of Sn have to be measured
from −600 to −300 mV to avoid any stripping current from other
possible metals such as Bi or Ni.
Typical SWASV results for Sn2+ sensing on the five BFEs with
their Bi crystal size ∼68, ∼39, ∼34, ∼30, and ∼27 nm are shown
in Fig. 6b as curves 1–5, respectively. Clearly, curves 3–5 show
the anodic stripping responses of Sn atoms which were previously
accumulated onto the BFEs at −0.8 V for 300 s from the testing
solution containing 35.4 ppm Sn2 under a magnetic stirring rate
of 100 rpm. On the other hand, the order of the SWASV curves
with respect to increasing their peak current density, centered
Fig. 6. The SWASV curves measured in 0.1 M citric acid for Bi deposits, (a) with-
out and (b) with Sn pre-deposition, prepared from the basic plating bath with (1)
−
2
−
2
0
0
.3 M CA + 0.3 M EDTA, at 1 mA cm , (2) 0.3 M CA + 0.2 M PEG, at 30 mA cm , (3)
−
2
.3 M CA + 0.3 M EDTA + 0.2 M PEG, at 30 mA cm , (4) 0.3 M CA + 0.3 M EDTA + 0.2 M
−
2
PEG + 4000 ppm gelatin, at 30 mA cm , and (5) 0.3 M CA + 0.2 M PEG + 4000 ppm
+
−2
gelatin, at 30 mA cm . The cathodic pre-deposition of Sn on BFEs is carried out at
−0.8 V for 300 s from a 0.1 M CA solution containing 35.4 ppm Sn2+. Curve 6 in (a) is
the SWASV response of Ni/Cu and curve 7 is the SWASV response of curve 4 in (b).

Y.-D. Tsai et al. / Electrochimica Acta 56 (2011) 7615–7621
7621
is close to 1, the concentration of Sn2+ from 0.15 to 35.4 ppm can
be precisely determined by the SWASV method. The above results
reveal that BFEs with small nanocrystals and a porous morphology
2+
will be a good electrochemical sensors for detecting trace Sn . The
optimization of BFEs for detecting trace Sn2+ and the other heavy
metal ions is going to be carried out very recently in this lab.
4. Conclusions
Based on the LSV analyses, the deposition behavior of Bi3+ ions
in the plating bath containing various combinations of complex
agents and additives can be understood. PEG and gelatin mainly
work as a leveling agent adsorbed on the surface of Bi deposits. From
the SEM images, PEG improves the adhesion and suppresses the for-
mation of dendrites for Bi deposits, resulting in a smoother film. On
the contrary, the formation of a compact shielding layer of gelatin
microspheres leads to localized hydrogen evolution on the deposit
and deeply cracked deposits with poor adhesion are obtained. The
binary PEG–gelatin additives avoid the above phenomenon and
favor to form sphere-like Bi nanocrystals with a porous morphology
−
2
at a high current density of deposition (30 mA cm ). The mor-
phology and crystal size (from 27 to 68 nm) of Bi deposits can be
controlled by varying the combination of CA, EDTA, PEG, and gelatin
as well as the current density. From the SWASV results, the deposit
with a smaller crystal size and a porous morphology exhibits the
excellent sensitive ability to Sn2+ with a limiting detection concen-
tration of 0.15 ppm.
Acknowledgments
The financial support of this work by the National Science Coun-
cil of ROC, Taiwan and the boost program of NTHU, is gratefully
acknowledged.
Fig. 7. The SWASV curves measured in 0.1 M citric acid for the Bi deposit plated from
the basic plating bath with 0.3 M CA + 0.3 M EDTA + 0.2 M PEG + 4000 ppm gelatin at
References
−
2
3
0 mA cm . (a) The cathodic pre-concentration of Sn on the BFEs is carried out at
0.8 V for 300 s from a 0.1 M CA solution containing (1) 0.03, (2) 0.15, (3) 3.5, (4)
[
[
1] X. Liu, P. Sun, S. Ren, L. Wen, Electrochem. Commun. 10 (2008) 136.
2] E. Hutton, B. Ogorevc, S. Hoevar, F. Weldon, M. Smyth, J. Wang, Electrochem.
Commun. 3 (2001) 707.
−
2+
8
.9, (5) 17.7, (6) 26.6, and (7) 35.4 ppm Sn . (b) The calibration plot for the SWASV
peak current against the Sn concentration.
2+
[3] K. Takamura, H. Shimamura, N. Koshiba, Thin film for package of alkaline bat-
tery and thin air battery using the same, Google Patents (2005).
[
[
4] F. Yang, K. Liu, C. Chien, P. Searson, Phys. Rev. Lett. 82 (1999) 3328.
5] F. Yang, K. Liu, K. Hong, D. Reich, P. Searson, C. Chien, Science 284 (1999)
around −480 mV, is: curve 3 (0.48 ␮A) < curve 5 (2.03 ␮A) < curve 4
1335.
(
13.08 ␮A) meanwhile no detective information is found on curves
[
6] B. Popov, R. White, G. Zheng, Corrosion 51 (1995) 429.
1
and 2. These results reveal that the sensing ability of BFEs is signif-
[7] R. Kachoosangi, C. Banks, X. Ji, R. Compton, Anal. Sci. 23 (2007) 283.
[8] F. Arduini, J.Q. Calvo, G. Palleschi, D. Moscone, A. Amine, TrAC-Trends Anal.
Chem. 29 (2010) 1295.
icantly influenced by the crystal size and morphology of Bi deposits.
In addition, the smaller nano-sized crystal and porous morphol-
ogy of a Bi deposit are, the better sensitivity for Sn2 detection can
be obtained, suggesting that the amount of Sn previously deposit-
ing on the Bi deposits depends on the active surface area of the Bi
deposits.
[
9] X. Duan, J. Yang, W. Zhu, X. Fan, C. Xiao, Mater. Lett. 61 (2007) 4341.
10] S. Jiang, Y. Huang, F. Luo, N. Du, C. Yan, Inorg. Chem. Commun. 6 (2003) 781.
[11] L. Li, Y. Zhang, G. Li, X. Wang, L. Zhang, Mater. Lett. 59 (2005) 1223.
+
[
[
[
[
12] D. Beutler, N. Giordano, Phys. Rev. B 38 (1988) 8.
13] D. Kim, S. Lee, J. Kim, G. Lee, Appl. Surf. Sci. 252 (2006) 3525.
14] S. Cho, Y. Kim, A. Freeman, G. Wong, J. Ketterson, L. Olafsen, I. Vurgaftman, J.
Meyer, C. Hoffman, Appl. Phys. Lett. 79 (2001) 3651.
15] Y.-D. Tsai, C.-C. Hu, C.-C. Lin, Electrochim. Acta 53 (2007) 2040.
16] Y.-D. Tsai, C.-C. Hu, J. Electrochem. Soc. 156 (2009) D58.
17] Y.-D. Tsai, C.-C. Hu, J. Electrochem. Soc. 156 (2009) D490.
[18] C.-C. Hu, Y.-D. Tsai, C.-C. Lin, G.-L. Lee, S.-W. Chen, T.-C. Lee, T.-C. Wen, J. Alloys
Compd. 472 (2009) 121.
19] C.-C. Hu, G.-L. Lee, J. Chin. Inst. Eng. 37 (2006) 589.
20] L. Bonou, M. Eyraud, R. Denoyel, Y. Massiani, Electrochim. Acta 47 (2002)
4139.
Since the BFE deposited at 30 mA cm−2 from the bath contain-
ing 0.3 M CA, 0.3 M EDTA, 0.2 M PEG, and 4000 ppm gelatin shows
the highest sensitivity in the testing solution, the sensing ability
of this deposit is further examined by changing the Sn2 concen-
tration in the testing solution. Typical results are shown in Fig. 7a
and the plot for the peak current density against the concentration
of Sn2+ is shown in Fig. 7b. Clearly, Fig. 7b displays a linear rela-
[
[
[
+
[
[
[
21] W. Dow, H. Huang, M. Yen, H. Huang, J. Electrochem. Soc. 152 (2005) C425.
tionship between the peak current density and the concentration
of Sn2+. This linear dependence of peak current density on the con-
[22] S. Wen, J. Szpunar, Electrochim. Acta 50 (2005) 2393.
23] T. Pal, A. Ganguly, D. Maity, Bull. Chem. Soc. Jpn. 60 (1987) 3001.
[24] S. Derkatch, L. Petrova, V. Izmailova, B. Tarasevitch, Colloids Surf. A 152 (1999)
89.
25] H. Thakkar, R. Murthy, ARS Pharmaceutica 46 (2005) 19.
[26] T. Kushibiki, H. Matsuoka, Y. Tabata, Biomacromolecules 5 (2004) 202.
27] F. Walsh, M. Herron, J. Phys. D: Appl. Phys. 24 (1991) 217.
28] L. Guo, P. Searson, Electrochim. Acta 55 (2010) 4086.
29] E. Sandnes, M. Williams, U. Bertocci, M. Vaudin, G. Stafford, Electrochim. Acta
52 (2007) 6221.
[
2
+
centration of Sn can be perfectly fitted by the following equation:
1
[
2
i = 0.36C + 0.9 (R = 0.991)
[
[
[
where i and C indicate the peak current (in ␮A) and Sn2+ concen-
2
tration (in ppm), respectively. Since the regression coefficient, R ,