A highly sensitive ascorbic acid sensor using a Ni-Pt electrode

Article abstract of DOI:10.1016/j.electacta.2011.08.084

An amperometric sensor that measured ascorbic acid by the oxidation of the ascorbic acid on a Ni-Pt electrode was fabricated. The Ni component of the Ni-Pt alloy played a crucial role as a modifier that developed an erinaceous surface, which enlarged the sensing area and increased the sensitivity of the electrode. The Pt₈₂Ni₁₈ electrode exhibited the best sensitivity of 333 μA cm^-2 mM^-1 for ascorbic acid sensing. This electrode was further tested for reproducibility of the sensitivity, endurance, and interference; it exhibited excellent performance compared with electrodes reported in the literature.

Full text of DOI:10.1016/j.electacta.2011.08.084

Electrochimica Acta 56 (2011) 9937–9945
Contents lists available at SciVerse ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
A highly sensitive ascorbic acid sensor using a Ni–Pt electrode
a,∗
b
a
c
Yu-Ching Weng , Yuan-Gee Lee , Yu-Lun Hsiao , Chiu-Yue Lin
a
Department Chemical Engineering, Feng Chia University, Taichung 40724, Taiwan
Department of Automation Engineering and Institute of Mechatronoptic Systems, Chien-Kuo Institute of Technology, Changhua 50094, Taiwan
Department Environmental Engineering and Science, Feng Chia University, Taichung 40724, Taiwan
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 17 June 2011
Received in revised form 20 August 2011
Accepted 22 August 2011
Available online 30 August 2011
An amperometric sensor that measured ascorbic acid by the oxidation of the ascorbic acid on a Ni–Pt elec-
trode was fabricated. The Ni component of the Ni–Pt alloy played a crucial role as a modifier that developed
an erinaceous surface, which enlarged the sensing area and increased the sensitivity of the electrode. The
Pt82Ni18 electrode exhibited the best sensitivity of 333 ␮A cm mM for ascorbic acid sensing. This elec-
trode was further tested for reproducibility of the sensitivity, endurance, and interference; it exhibited
excellent performance compared with electrodes reported in the literature.
−2
−1
Keywords:
Ni–Pt
©
2011 Elsevier Ltd. All rights reserved.
Sensing mechanism
Erinaceous surface
Sensitivity
Reliability
1
. Introduction
potentiometric sensors are often limited by slow response times
and ionic noise from the solution [8].
Albert Szent-Gyorgyi isolated ascorbic acid (H A) from the
adrenal gland of a cow and named it vitamin C. An insufficient sup-
Amperometric sensors can be made to overcome the disadvan-
tages of other sensors. An amperometric sensor is usually designed
as a laminate structure with a metal film on top of an active layer,
which is, in turn, on top of a substrate. The metal film is pro-
portionally sensitive to specific chemicals. Amperometric sensors
respond to H A quickly and can detect a wide range of H A con-
2
ply of H A in the human body causes symptoms of scurvy [1], but
2
excessive H A results in stomach convulsions [2]. H A promotes
2
2
healthy cell development, normal tissue growth, and the healing
of injuries and burns [3]. It helps in the synthesis of collagen, an
important component of ligaments, blood vessels, cartilage, bones,
and tendons [4]. Physiologists are devoting increasing attention to
2
2
centrations [9–13]. Many amperometric sensors contain modifier
metals in their metal films to enhance the electrode response. In
the quantitative measurement of H A.
this study, we used platinum as the H A-sensitive metal matrix
2
2
The direct measurement of the amount of ascorbic acid in the
human body is difficult because ascorbic acid is water soluble,
acts as an antioxidant, and is not stored in the body. Various
[14–16] and nickel as the modifier metal because nickel is chemi-
cally similar to platinum. The Pt-lean (Pt content less than 50%) and
Pt-rich (Pt content greater than 50%) alloys were employed to sense
researchers have developed H A sensors with specific functions
H A simultaneously. However, the Pt-lean alloys are less sensitive
2
2
and advantages. Sensors based on ion-sensitive field-effect tran-
sistors (ISFETs) are highly stable and respond quickly, but IFSET
than Pt-rich alloys for H A oxidation, which is not presented in this
2
paper. We further confirmed the modifying theorem in which the
rich component, Pt, is the sensing metal and the lean component,
Ni, serves as a modifier to enhance the sensitivity by intensifying
the electrode’s specific area. After a series of electrochemical tests,
we found that the Ni–Pt alloys were more sensitive than either Ni
or Pt in isolation.
lithography is expensive [5]. Optical H A sensors can detect H A in
2
2
a reaction mixture, but various environmental factors, e.g., chem-
ical disturbances, temperature, and optical conditions, can limit
the effectiveness of such sensors [6]. Conductometric sensors can
be operated with ease but often exhibit unsatisfactory sensitiv-
ity, selectivity, and accuracy levels [7]. Potentiometric sensors can
be effective even if they use miniature working areas; however,
2. Experimental
2.1. Fabrication of the Ni–Pt electrode
∗ Corresponding author. Tel.: +886 4 24517250x3689; fax: +886 4 24510890.
E-mail address: ycweng@fcu.edu.tw (Y.-C. Weng).
An alumina board was cut into slices, each of which measured
3.5 cm × 1.5 cm × 0.1 cm. These slices were cleaned ultrasonically
0
013-4686/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2011.08.084

9
938
Y.-C. Weng et al. / Electrochimica Acta 56 (2011) 9937–9945
Au/Al
2
O
3
3. Results and discussion
3
3
.1. Microstructure investigation of the Ni–Pt electrode
0.3 cm
.1.1. Morphology investigation
The portion of the Ni–Pt electrode with a Ni matrix contained
agglomerated spinodal crystals whereas the portion with a Pt
matrix was characterized by coarse grains. The local morphology
of the Ni–Pt electrode varied according to the local matrix. Fig. 1(a)
shows an erinaceous surface that was observed on a Ni crystal.
Ni film was particularly notable on the embedded needles. The
three images of Fig. 1(b)–(d) show how Pt caused the spinodal sur-
faces to mix with the particle phases. These images indicate that Pt
atoms affected crystallization as they blended into the Ni crystal lat-
tices. Fig. 1(e) shows a surface for which all spinodal crystals were
replaced with coarse particles. The high sensitivity of the surface
was determined by two conditions: a high specific area, i.e., high
surface area per unit weight of crystal, and a sensitive preferred
crystal orientation on the electrode [17]. Fig. 1(a) shows a spinodal
surface with a large specific area; Fig. 1(e) shows a coarse surface
littered with agglomerate particles. With regard to the preferred
crystal orientation, we discuss in Section 3.1.2 how crystallization
can be identified by diffraction patterns.
2 3
Al O substrate
1.4 cm
Scheme 1. The screen printing pattern for the conductive Au paste.
in an acetone bath for 20 min to remove the organic coating. The
cleaned slices were then rinsed with deionized (DI) water and dried
in an oven at 100 C for 20 min. A “doctor blade” screen-printing
process [14] was used to deposit conductive gold paste (Heraeus
C-5754B, 87.5%) onto each aluminum substrate to form the pattern
shown in Scheme 1. Each as-printed substrate was maintained at
◦
◦
1
50 C for 40 min to evaporate the solvent in the paste and was
◦
then heated in a furnace (CHEMIST DF20) at 850 C for 40 min
to sinter the gold layer. Each as-sintered substrate was soaked
in an acetone bath for 30 min and then cleaned in DI water for
3
0 min.
The electrode layers were prepared by electrodeposition. A
3.1.2. Composition-induced phase transformation in the Ni–Pt
alloys
reference electrode made of Ag/AgCl (KCl sat’d) was configured
to maintain a standard applied potential. To prevent inhomo-
geneous deposition, a stirring plate was used to agitate all the
deposition solutions. For the pure Pt layer, a solution of 0.01 M
H PtCl ·6H O (dihydrogen hexachloroplatinate (IV) hexahydrate
An alloy that incorporates Ni into a Pt matrix is expected to
exhibit a porous morphology. Fig. 2 shows a series of X-ray diffrac-
tion patterns for the Ni–Pt alloys. The intensities of the Au(1 1 1)
peaks in both pure Ni and pure Pt are smaller than those in Pt61Ni39
,
2
6
2
Pt70Ni30, and Pt82Ni18 because the Au(1 1 1) signal originates from
the sub-layer. A lower Au(1 1 1) intensity reflects a smaller exposed
Au sub-layer area. A small exposed sub-layer area results in a large
exposed deposited layer. We concluded that the porous morpholo-
gies of the alloyed compositions Pt61Ni39, Pt70Ni30, and Pt82Ni18
tended to reflect signals from the substrates.
9
9.9%, UR) was used at an applied potential of −100 mV in a
potentiostat (CHI 824B). For the pure Ni layer, a solution of 1 M
NiCl ·6H O (nickel chloride hexahydrate, SHOWA, 96%) was used
2
2
at an applied potential −800 mV. The deposition rate between Ni
and Pt strongly depends on the applied potential. We therefore
adjusted the alloy composition by increasing the deposition rate
of Ni and reducing the deposition rate of Pt. The distinct applied
potentials were applied to differentiate the composition of the
Ni–Pt alloys. A solution of 0.01 M H PtCl ·6H O was blended with
The intermetallic compounds (IMCs) of the Ni–Pt alloys exerted
critical effects on the microstructural variations of the electrodes.
Fig. 2 shows that a Pt(1 1 1) peak was present for all compositions
except pure Ni. Pure Ni does not show this peak, but materials with
some Pt show a peak proportional to the amount of Pt. This result
indicates that the Pt component enhanced the Pt(1 1 1) preferred
orientation. Only pure Ni displayed the Ni(2 0 0) peak. This result
suggests that the Ni lattice was reconstructed as soon as Pt atoms
infiltrated the Ni matrix. We further confirmed the result that only
two IMCs, Ni Pt and NiPt , contributed to the intermediate com-
2
6
2
a solution of 1 M NiCl ·6H O, and potentials of −200 mV, −400 mV,
2
2
and −600 mV were applied. All depositions were performed for
1
0 min.
2.2. Instruments for the investigation of the electrode
3
3
positions Pt61Ni39, Pt70Ni30, and Pt82Ni18; details are provided in
Appendix A.
The morphology of each electrode was investigated with a
scanning electron microscope (Hitachi 3000-H). Electrode com-
positions were measured with an energy-dispersive spectrum
analyzer (Noran Voyager 2.0). The crystallization of the Ni–Pt alloys
was measured with a thin-film X-ray diffractometer (XRD) (Bruker
D8-SSS).
3
3
.2. The electrochemical reaction on the electrode
.2.1. Cyclic voltammetry
The results of the cyclic voltammetry (CV) tests were compli-
cated by faradic and non-faradic currents that contributed to the
peak shifts in this study. The following paragraphs describe the
results of CV tests on the Pt, Ni–Pt and Ni electrodes with both
2.3. Preparation of the test solution
blank and H A-containing solutions.
The Pt electrode was found to be sensitive to ascorbic acid.
The CV measurement of Pt subjected to H A is shown in Fig. 3.
In the blank solution, the response current in the forward scan
increased appreciably at −0.68 V and −0.62 V, as shown by the
dashed line. The first peak was related to the desorption pro-
cess for the weakly bound hydrogen ions on the electrode; the
second peak showed the desorption process for hydrogen atoms
strongly bound to platinum [18]. Pitara and co-workers investi-
gated the preferred orientations of the adatom peaks; they found
2
Ascorbic acid (C H O , l-(+)-ascorbic acid 99%, Acros) was
diluted to serve as a stock solution for the subsequent blend-
ing. The alkaline test solution was a blank solvent that contained
6
8
6
2
0
.1 M KOH (85%, Showa). The interference reagents for the
H A test were anhydrous oxalic acid (99%, H C O , Showa),
2
2
2
4
sucrose (>99.9%, C12H22O11
, Showa), sodium benzoate (99%,
C H5COONa, Showa), l-(+)-tartaric acid (99%, C H O , Showa), d-
6
4
6
6
(
−)-fructose (99%, C H O , Acros), and anhydrous citric acid (99%,
6
12
6
C H (OH)(COOH) , Showa).
3
4
3

Y.-C. Weng et al. / Electrochimica Acta 56 (2011) 9937–9945
9939
Fig. 1. SEM images of the Pt–Ni alloy electrodes. (a) Ni; (b) Pt61Ni39; (c) Pt70Ni30; (d) Pt82Ni18; (e) Pt on Au/Al2O3 substrates.
that the first peak usually occurs on the Pt(1 1 0) orientation and
that the second peak preferred the Pt(1 0 0) orientation [19]. Both
of these responses occurred on the polycrystalline Pt electrode.
No obvious reduction peak was observed in the reverse scan to
couple with the adsorption process, which suggests that the elec-
trode might have been poisoned and that the reduction current
might have been inhibited. Gasteiger et al. researched the oxida-
tion of CO/H2 mixtures on Pt–Ru alloys and found that even trace
amounts of CO negatively shifted the potential [20]. Igarashi et al.
studied electro-oxidation on Pt catalysts and concluded that the
type of CO bonding on Pt sites depended on the CO coverage [21].
These results imply that the CO exclusively occupied the active sites
of the electrode and repelled the hydrogen ions [22]. The reduc-
tion peak appeared only when the applied voltage was extended
to negative potentials beyond the measured range of this study.
When the forward scan voltage was increased, a plateau appeared
at −0.1 V because of Pt oxidation current on the electrode. In the
reverse scan, a conspicuous peak appeared at −0.28 V because of
PtOx reduction. The voltammogram for the H A-containing solu-
2
tion showed a distinct response, as shown by the solid line in
Fig. 3. Two broad peaks for the adsorption/desorption processes
still appeared at −0.65 V and −0.45 V, but they shifted forward.
Two waveform peaks were observed at 0.14 V and 0.35 V. The
first peak, at 0.14 V, represented a series of oxidation processes
with the product 2,3-diketogulonic acid (DKG); the second peak
showed the oxidation of the DKG. Either the first or the second
peak showed a two-electron process, which means the oxidation
was not a single-electron reaction. These two peaks showed broad
waveforms and high current intensities. Oxidation of Pt electrode
itself also occurred simultaneously in this potential region. Thus,

9
940
Y.-C. Weng et al. / Electrochimica Acta 56 (2011) 9937–9945
1.5
1.2
0.9
0.6
0.3
Ni
Pt Ni
39
61
without H A
2
Pt Ni
30
70
5
.68 mM H A
2
Pt Ni
8
2
18
0.0
-0.3
Pt
-0.6
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
3
0
35
40
45
50
55
60
2
θ
E vs (Ag/AgCl)/V
Fig. 2. X-ray diffraction patterns of the PtxNiy electrodes.
Fig. 4. Cyclic voltammetric behaviors of a Ni electrode (- - -) without H2A and (—)
with 5.68 mM H2A in 0.1 M KOH solution.
a small, single reduction peak appeared at 0.0 V in the reverse
scan, which represented the reduction of PtOx on the electrode
In contrast to the blank solution reaction, the reduction peak at
[
23].
Fig. 4 shows the CV curves for a Ni electrode in both a blank
0
.36 V disappeared from Fig. 4 when the test solution contained
H A. This process is shown by the solid line in Fig. 4. When the H A
2
2
and an H A-containing solution. In the blank solution, the response
2
molecule approached the Ni electrode, it adsorbed on the surface
current increased appreciably at −0.03 V, as indicated by the dashed
line. This increase was due to the electrochemical reaction between
the Ni and the hydroxyl ion; ␣-Ni(OH)2 and ␤-Ni(OH)2 remained
on the electrode [24]. The reaction can be expressed as follows:
of Ni. The H A molecule was then electro-activated to hydrate and
2
to release hydrogen ions; therefore, the zeta potential was changed
and the local pH value near the electrode was reduced. The decrease
in pH inhibited the Ni(OH)2 oxidation as described in Eq. (2), but it
still allowed a series of oxidations to contribute a peak at 0.04 V as
described by:
Ni + 2OH− → (␣, ␤) − Ni(OH) + 2e−
(1)
2
When the applied voltage was increased to 0.46 V, ␤-type and
H A → HA + H+ + e−
−
(4a)
(4b)
␥-type Ni peroxide formed immediately. This phase transformation
2
was accompanied by an electron transfer, which occurred on the
electrode as expressed in the following formula [14]:
HA− → A2− + H+ + e−
−
2−
−
2−
represent C H7O6 and C H O , respec-
6
where HA and A
6
6
6
Ni(OH)2 → (␥, ␤) − NiOOH + H+ + e−
(2)
tively.
Fig. 5 shows the CV tests for the Pt82Ni18 electrode in both blank
and H A-containing solutions. The Ni–Pt alloys showed consider-
Furthermore, considerable amounts of ␥-type Ni oxide formed
by oxidation of ␥-NiOOH. This process can be described by the
following expression [25]:
2
able voltage shifts in the CV test with the H A-containing solution.
2
When the Pt82Ni18 electrode was scanned at positive potentials
␥
− NiOOH → NiO + H+ + e−
(3)
2
above −0.51 V in the blank solution, the formation of Pt Ni O
x
y
z
occurred. The oxidation of PtwNi
(OH) took place at poten-
2
The formation of NiOOH and NiO2 contributed to the peak at
.46 V. The reduction peak appeared at 0.36 V in the reverse scan.
1−w
tials greater than 0.45 V whereas the reduction of PtwNi
(OOH)
0
1−w
0
0
0
0
0
.4
.3
.2
.1
.0
2
1
5
.68 mM H A
2
0
without H A
2
-1
-
-
-
0.1
0.2
0.3
-
2
electrolyte alone
-3
5
.68 * 10 M H A
2
-3
-
0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
E vs (Ag/AgCl)/V
E vs (Ag/AgCl)/V
Fig. 3. Cyclic voltammetric behaviors of a Pt electrode (- - -) without H2A and (—)
Fig. 5. Cyclic voltammetric behavior of a Pt82Ni18 electrode (- - -) without H2A and
with 5.68 mM H2A in 0.1 M KOH solution.
(—) with 5.68 mM H2A in 0.1 M KOH solution.

Y.-C. Weng et al. / Electrochimica Acta 56 (2011) 9937–9945
9941
1
1
0
0
0
0
0
.2
.0
.8
.6
.4
.2
.0
1.2
a
5.11 mM
0.9
0.6
0.3
0.0
Region II
3.98 mM
Region I
2.84 mM
1.73 mM
0.57 mM
-
0.2
-
0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
0
1000
2000
3000
E vs (Ag/AgCl)/V
t/s
Fig. 6. The polarization curve of a Pt82Ni18 electrode with 5.68 mM H2A in 0.1 M
KOH solution.
b 1.2
commenced at 0.5 V in the reverse scan. A large PtxNiyOz reduc-
1.0
0.8
0.6
0.4
0.2
tion peak appeared at −0.36 V. In the presence of H A, the peaks
2
that correspond to hydrogen adsorption and desorption appear at
potentials between −0.7 and −0.4 V. The two peaks at 0.11 V and
0.32 V in the forward scan were related to the initial oxidation of
the H A and to its further oxidation, respectively. The oxidation of
2
the Pt82Ni18 electrode also occurred simultaneously in the forward
scan. Therefore, a distinct reduction peak of PtxNiyOz appeared at
−
0.02 V in the reverse scan. We concluded that the Ni–Pt alloy per-
mitted the H A electrochemical reaction. The results suggested that
2
Pt alloyed with Ni possesses an outstanding sensitivity for H A,
which benefits from particular microstructural variations of the
Ni–Pt alloys, as discussed in Sections 3.1.1 and 3.1.2.
2
0.0
0
1
2
3
4
5
6
7
3.2.2. Polarization curves for applied potentials
c/mM
The applied potentials were determined using polarization
curves. Fig. 6 shows the polarization curve by which both the
polarizing current and the applied potential were determined.
c
2.0
Ni
The response current for an H A-containing solution exhibited
2
Pt Ni
two stepwise variations that are shown with hollow symbols in
Fig. 6. The first step, from −0.1 V to 0.25 V, contained a ramp
and a flat region; this behavior indicated kinetics-controlled and
diffusion-controlled processes. For the kinetics-controlled process,
the current was increased because it related to the oxidation rate
61
39
1.5
1.0
0.5
0.0
Pt70Ni30
Pt Ni
82
18
Pt
of H A. The characteristics of the kinetics-controlled process pro-
2
duced a rush current in a high concentration of H A. This rush
2
current tended to prolong the disturbance time and prevent a sta-
ble response; therefore, the applied potential had to be excluded
for the kinetics-controlled process. In the flat region, labeled region
I, the current reached a constant value that was independent of
the reaction rate because of a diffusion-controlled process. From
0
1
2
3
4
5
6
7
0
.25 V to 0.50 V, the secondary oxidation of H A produced another
2
flat region, labeled region II. A higher voltage applied in region
II induced unnecessary responses from interferences, which dis-
turbed the measurement process. Thus, the applied potential data
were taken from range I, where the mathematical average potential
was 0.2 V.
c/mM
Fig. 7. (a) Amperometric response of a Pt electrode in 0.1 M KOH solution at an
applied potential of 0.2 V with a successive additions of H2A in the concentration
range of 0.57–5.68 mM. (b) The calibration curves fitted with a linear regression.
(
c) The calibration curves of the PtxNiy alloy electrodes in 0.1 M KOH solution at
an applied potential of 0.2 V with successive additions of H2A in the concentration
range of 0.57–5.68 mM. Symbols: (ꢀ) Ni/Au/Al2O3; (᭹) Pt61Ni39; (ꢁ) Pt70Ni30; (ꢂ)
Pt82Ni18; (ꢀ) Pt/Au/Al2O3.
3
.2.3. Chronoamperometric tests
H A concentration calibration curves were constructed for all
2
Ni–Pt electrodes. Fig. 7(a) shows the response current for the Pt
electrode in 0.1 M KOH solution with a fixed applied potential of
0.2 V. The response current increased stepwise for each 0.57 mM
increase in H A. With the increase in the H A concentration, the
2
2
noise increased because the large response current attracted stray

9
942
Y.-C. Weng et al. / Electrochimica Acta 56 (2011) 9937–9945
Table 1
Various types of amperometric sensors.
Electrode
Electrolyte
Linear range (M)
Sensitivity
Detection limit
(mM)
Ref
−
2
−1
(
␮A cm mM )
−
−
6
−5
−4
120a
406
GCE modified with ferricyanide-doped Tosflex
GCE modified with vanadium oxide
Phosphate buffer (pH = 5)
Britton–Robinson (pH = 8.06)
1 × 10 –5 × 10
–
–
[26]
[27]
8
4 × 10 –1 × 10
(
VO(OC3H7)3) polypropylene carbonate
SPCE modified with drop-casting of
dodecylbenzene sulfonic acid (DBSA)-doped
polyaniline nanoparticles
Phosphate buffer (pH = 6.8)
5 × 10−4–8 × 10−3
10.75
8.3 × 10−3
[28]
AGCE and AuE modified with ascorbate
oxidase (ASOD) micelle membrane
GCE modified with electrodeposited films
derived from dihydroxy-benzaldehyde
GCE modified with
Phosphate buffer (pH = 5.5)
Phosphate buffer (pH = 7)
0.1 M H2SO4
5 × 10−6–4 × 10−4
2 × 10−3–3 × 10−3
1 × 10−5–9.8 × 10−4
10.85 (AGCE)
–
[10]
[12]
[29]
0.44 (2,5-DHB)
687
0.16 (2,5-DHB)
5.5 × 10−3
octacyanomolybdate-doped
poly(4-vinylpyridine)
−
3
Screen-printing ruthenium dioxide electrode
SPCE modified with MWNTs/PVI-Os multilayer
films
Graphite electrode modified with polypyrrole
nanowire
Phosphate buffer (pH = 7.4)
Phosphate buffer (pH = 7)
0–4 × 10
2.67
12.5
1.1
3 × 10
[30]
[31]
5 × 10 7–7 × 10−5
−
a
−4
0.1 M phosphates
5 × 10−4–2 × 10−2
46.51
4.65 × 10−2
[32]
−
7
−4
460a
233
−5
GCE modified with poly(glutamic acid)
Boron-doped diamond electrode modified
with ferrocyanide-trapped PDMA film
Thermally pre-treated iron electrodes
modified with polypyrrole and ferrocyanide
Phosphate buffer (pH = 5)
Phosphate buffer (pH = 7)
1.2 × 10 –2.5 × 10
4 × 10
[33]
[34]
−
6
−5
1 × 10 –3 × 10
–
5 × 10− –9 × 10−3
4
120
0.15
[35]
Phosphate buffer (pH = 7)
ions Fe(CN)63 (FCN)
−
Carbon ceramic electrodes modified with
terbium hexacyanoferrate
Hexacyanoferrate (II/III) and poly-d-lysine was
immobilized on a gold electrode modified
with cysteamine
0.5 M KCl (pH = 7)
TRIS buffer (pH = 7)
5 × 10−7–1 × 10−4
138.8a
62.9
2 × 10−4
[36]
[37]
1 × 10−6–9.1 × 10−4
2.2 × 10−3
Carbon paste electrode modified with
0.5 M KCl (pH = 7)
7.9 × 10−4–6.7 × 10−3
0–6.2 × 10−
4.53
23.3
190
–
[38]
[37]
[39]
[13]
(
silica/aniline, Si–An)/(Congo red, CR)
3
3.1 × 10−4
–
GCE modified with Ni(Me2(CH3CO)2
tetraenoN4) complex
Nickel electrode modified with polyaniline
Phosphate buffer (pH = 6.6)
0.1 M H2SO4
5 × 10− –3.5 × 10−2
3
(
PANI)
The electrochemical copolymerization of
,4-dihydroxybenzoic acid and aniline at
Phosphate buffer (pH = 6.6)
1 × 10−4–1 × 10−2
240a
5 × 10−2
3
microdisk gold electrode
Carbon paste electrode modified with
NH3–NH4Cl buffer (pH = 10.0)
5 × 10−7–1 × 10−5
65.5a
5 × 10−4
[40]
(
␤-cyclodextrin)–ferrocene inclusion
complex
Pt foil
−
4
−3
−3
−3
0.1 M KOH
0.1 M KOH
0.1 M KOH
5.68 × 10 –5.68 × 10
178
169
333
–
–
–
[43]
[43]
–
−
4
Ni foil
Ni–Pt alloys (this study)
5.68 × 10 –5.68 × 10
−4
5.68 × 10 –5.68 × 10
a
−1
Geometric area is unknown (unit = ␮A mM ).
ions over the active site of the electrode. This dependence was re-
plotted and fitted with a linear relationship to obtain the electrode’s
sensitivity, 177 ␮A cm mM , as shown in Fig. 7(b). Fig. 7(c)
shows calibration curves for the Ni–Pt electrodes. The sensitivi-
ties follow the inequality Pt82Ni18 > Pt70Ni30 > Pt61Ni39 > Pt > Ni. The
3.3. Reliability of the Ni–Pt sensor
3.3.1. The effect of pH value on the sensitivity
The pH values of the analyzed solutions dominated the electro-
chemical kinetics; either an alkaline solution or an acidic solution
could work as an analyzed solution. In acidic solutions, the hydro-
−2
−1
mid-composition Ni–Pt alloys – Pt82Ni18, Pt70Ni30, and Pt61Ni39
–
exhibited higher sensitivities than the pure metals, Ni and Pt. The
sensitivity of pure Ni was so low that it was almost negligible. We
propose that, in the bimetallic sensors presented here, Pt was the
gen ions of H A were difficult to decompose because the high
concentration of hydrogen ions in the solutions created a co-ion
2
effect. The H A might have directly oxidized with the release of two
2
main component that sensed H A, and Ni was a relatively inert
component. The Ni did not exhibit high sensitivity, but enhanced
the sensor by varying its microstructure, as discussed in Sections
electrons, as described in Eqs. (4a) and (4b); these intermediates
might have further hydrolyzed to become hydrated dehydroascor-
bic acid in an acidic solution. Inversely, the hydrogen ion is apt
to dissociate in an alkaline solution at pH = 10–11. Ruiz and co-
workers researched the oxidation mechanism in alkaline solution
and found a reaction order of “−2.” In our study, as the alkalin-
2
3
.1.1 and 3.1.2. The alloying of the Ni component in the Pt matrix
induced a preferred sensitive orientation with a spinodal morphol-
ogy that enlarged the sensor’s specific area and provided more
sensing sites. Thus, the sensitivity levels of the mid-composition
Ni–Pt alloys – Pt82Ni18, Pt70Ni30, and Pt61Ni39 – were notably high.
The Ni–Pt ascorbic acid sensor performed better than other
amperometric sensors. Table 1 lists the sensitivity levels of various
ascorbic acid sensors. Excluding sensors with carbon-based elec-
trodes, the Ni–Pt electrode (described in this study) showed a very
ity was increased to pH 12–13, the H A dissociated two hydrogen
2
ions to produce DKG. The reaction order was “−1” because the
hydroxide ion in the alkaline solution tended to neutralize with
the decomposed hydrogen ions. The reactions described in Eqs.
(4a) and (4b) went forward according to LeChatelier’s principle.
The alkaline solution was capable of producing high sensitivity and
was therefore adopted in this study.
−2
−1
high sensitivity of 333 ␮A cm mM .

Y.-C. Weng et al. / Electrochimica Acta 56 (2011) 9937–9945
9943
25
20
15
10
5
0
acid measurements. Because its normal concentration in the human
−
1
−1
−1
body ranges from 3.6 mg dl
to 8.3 mg dl , a 6.4 mg dl stock
solution of saturated uric acid was blended into the analyzed solu-
tion. We observed almost no noise in the response current, as
shown in Fig. 8. Fig. 9 shows that very little noise was observed
when the other reagents were tested in a similar manner.
4. Conclusion
0.2 mM
0.4 mM
2
.4 mM
The Ni–Pt electrode showed high sensitivity, endurance, and
reliability in this study. We found that the sensing mechanism of the
Ni–Pt electrode depended not only on the oxidation of the ascor-
bic acid in the solution but also on the crystallization of the Ni–Pt
itself. In addition, most of the oxidation resulted from charge trans-
fer from Pt rather than from Ni. Although both Pt and Ni were able to
sense H A, the Ni–Pt alloy exhibited higher sensitivity than either
0
40
80
120
160
2
metal used alone. We attribute the Ni component’s enhanced sensi-
tivity to the alloy’s erinaceous surface that enlarged the electrode’s
sensing area.
t/s
Fig. 8. Response current vs. elapsed time for different blending concentrations of
uric acid, 0.2 mM, 0.4 mM, and 2.4 mM, in the 0.1 M KOH solution.
Acknowledgements
3
.3.2. Response time
The response time followed a diffusion-controlled process
We gratefully acknowledge support from the National Science
Council of the Republic of China (NSC 99-2221-E-035-093) and
from Feng Chia University (Grant No. 10G27101).
caused by the fast redox reaction on the electrode. The reaction rate
was determined by the diffusion of the reactant. For dilute solu-
tions, the gradients of the H A concentration near the electrode
2
were low; therefore, the driving forces were weak, and the H A
Appendix A.
2
molecules required more time to reach the electrode. In contrast,
solutions with high H A concentration gradients also had strong
driving forces and produced faster response times.
Few articles in the literature have discussed the Ni–Pt phase dia-
gram. Dahmani et al. have published a qualitative series of binary
phase diagrams, including a Ni–Pt system [25]. Unfortunately, their
Ni–Pt phase diagram is spanned by a complete series of random ␣-
solid solutions and, because of the long duration of Ni–Pt phase
separation, the diagram does not show low temperatures. Theoret-
ical microstructures for Ni–Pt alloys, including Ni, Ni3Pt, NiPt, NiPt3,
and Pt, have been established. Okamoto has reported the Ni–Pt
crystal structure data shown in Table A1 [41]. However, no infor-
mation on the binary Ni–Pt alloy was found in the JCPDS database.
We therefore developed theoretical lattices and their diffraction
2
3.3.3. Stability test
The sensitivity exhibited a stable value with a small deviation.
The as-prepared electrode was tested for 67 days; the standard
deviation for the sensitivity was 3.3%. We considered the sensor
to have good endurance.
3.3.4. Selectivity test
Chemical noise significantly affected the sensitivity of the Ni–Pt
patterns to describe Ni–Pt alloy compositions.
electrode. We tested general disturbances from medical auxiliary
reagents, including uric acid, glucose, fructose, sucrose, tartaric
acid, citric acid, oxalic acid, and sodium benzoate. A series of reagent
blending processes were performed to test the electrode’s selectiv-
ity. For example, uric acid is often a disturbance source for ascorbic
The Ni crystal has been measured: its atomic radius is 1.62 A˚ , its
lattice parameter is 2.62 A˚ , and its space group is Fm3m [42]. The
software package CaRIne was employed to fit the Ni data. The Ni
atomic coordination arrangement and its theoretical XRD pattern
are shown in Fig. A1. This theoretical pattern was confirmed by the
JCPDS data. IMC lattices for Ni Pt, NiPt, and NiPt must be stretched
3
3
to contain the larger Pt atom, whose atomic radius is 1.82 A˚ larger
350
300
250
200
150
100
(
i.e., 10.4% larger) than that of Ni. The IMC lattice was, therefore,
fitted with the Pt lattice parameter of 3.94 A˚ . Figs. A2–A4 show the
IMCs’ atomic coordination arrangements and their theoretical XRD
patterns. The Pt XRD pattern was constructed by applying the data
in JCPDS, as shown in Fig. A5. Because the peak at ca. 40 disap-
peared in the diffraction pattern of NiPt, this alloy was disregarded
in this research.
Only part of the intermediate phases between Ni and Pt
are presented in this study. Fig. A6 presents a Ni–Pt binary
◦
Table A1
50
Ni–Pt crystal structure data modeled after the results of Okamoto [41].
Phase
Pearson symbol
Space group
Superlattice
structure
0
ascorbic glucose fructose sucrose tartaric citric oxalic sodium
--
acid
acid
acid
acid benzote
(Ni, Pt)
Ni3Pt
NiPt
cF4
cP4
tP4
cP4
Fm3m
Pm3m
P4/mmm
Pm3m
A1
L12
L10
L12
Interference
NiPt3
Fig. 9. Selectivity of the Pt82Ni18 electrode in 0.1 M KOH solution.

9
944
Y.-C. Weng et al. / Electrochimica Acta 56 (2011) 9937–9945
Intensity (%)
1
, 1, 1
c
Intensity (%)
00
1, 1, 1
39. 76, 100. 0)
(
44. 51, 100. 0)
1
00
Pt
a
(
1
9
0
0
0
0
0
0
0
0
0
Ni
z
b
x
c
9
0
y
y
x
Ni
8
z
b
80
a
7
6
5
4
3
2
1
7
6
0
0
2
, 0, 0
2, 0, 0
( 46. 24, 46. 9)
(
51. 86, 44. 7)
50
1
, 0, 0
40 ( 22. 65, 36. 2)
1, 1, 0
( 32. 24, 30. 0)
3
, 1, 1
92. 96, 23. 4)
2, 2, 0
30
20
(
(
76. 40, 21. 6)
2, 1, 0
( 52. 08, 15. 1)
2
, 1, 1
( 57. 49, 11. 1)
2
, 2, 2
1
0
0
( 98. 47, 6. 8)
2
q ( ¢
2
q (¢
0
2
0
25
30
35
40
45
50
55
60
0
10
20
30
40
50
60
70
80
90
100 110 120
Fig. A4. The theoretical XRD pattern of NiPt3 and its atomic arrangement.
Fig. A1. The theoretical XRD pattern of Ni and its atomic arrangement.
Intensity (%)
1
, 1, 1
Pt
( 39. 76, 100. 0)
Intensity (%)
00
100
90
80
70
60
50
40
1
, 1, 1
b
(
39. 76, 100. 0)
1
Ni
b
y
x
9
0
0
a
y
z
Pt
c
z
8
x
a
7
6
0
0
2
, 0, 0
c
(
46. 24, 47. 6)
2
, 0, 0
5
4
3
2
1
0
0
0
0
0
( 46. 24, 46. 9)
1
, 0, 0
3
2
1
0
0
0
( 22. 65, 36. 2)
1, 1, 0
32. 24, 30. 0)
(
2
, 1, 0
2, 1, 1
(
52. 09, 15. 1)
(
57. 50, 11. 1)
2 q (¢
60
0
2
0
25
30
35
40
45
50
55
2
q ( ¢
0
2
0
25
30
35
40
45
50
55
60
Fig. A5. The theoretical XRD pattern of pure Pt and its atomic arrangement.
Fig. A2. The theoretical XRD pattern of Ni3Pt and its atomic arrangement.
Pt70Ni30, and Pt82Ni18 exhibited grains that were finer than
those in Fig. 1(e) because of the specific coarse-grain morphol-
ogy of the electrodeposited Pt crystals. The phase of Pt was
also excluded in these intermediate compositions. Therefore,
the left IMCs, i.e., Ni3Pt and NiPt3, contributed the interme-
phase diagram. Five thermodynamically stable phases exist: Ni,
Ni Pt, NiPt, NiPt , and Pt; the theoretical XRD patterns present
3
3
reconstructed microstructures of these phases. The particular
◦
peak at ca. 39.8 in Fig. 2 appeared in most of the diffrac-
diate compositions Pt61Ni39
, Pt70Ni30, and Pt82Ni18. Ni was
tion patterns except that of Ni; however, it did not appear in
Fig. A3 for the NiPt phase. The NiPt phase was excluded in
regarded as a modifier; it characterizes the preferred crystal-
lizations in this study. However, the lattice parameters of Ni Pt
3
the alloys containing Pt. The intermediate compositions Pt61Ni39
,
Intensity (%)
700
1
, 0, 0
(
22. 66, 100. 0)
c
100
600
(Ni,Pt)
Experimental composition
9
0
0
0
0
0
0
0
0
0
Ni
z
500
400
300
200
8
x
y
1
, 0, 1
7
6
5
4
3
2
1
( 34. 19, 65. 7)
a
b
Pt
Ni Pt
NiPt
NiPt3
3
0
, 0, 1
1, 1, 0
1
, 1, 1
25. 26, 37. (8 3) 2. 27, 38. 8)
( 41. 41, 36. 8)
(
2
, 1, 1
2
, 1, 0
2
, 0, 1( 58. 78, 22. 2)
2
, 0, 0
1, 0, 2
100
0
( 52. 13, 17. 1)
(
53. 45, 15. 6)
(
46. 28, 12. 8)
( 57. 29, 12. 2)
0
, 0, 2
(
51. 86, 4. 3)
2
q ( ¢
0
10
20
30
40
50
60
70
80
90
100
Pt
0
Ni
2
0
25
30
35
40
45
50
55
60
Atomic Percent Platinum (%)
Fig. A3. The theoretical XRD pattern of NiPt and its atomic arrangement.
Fig. A6. The Ni–Pt binary phase diagram modeled after the results of Okamoto [41].

Y.-C. Weng et al. / Electrochimica Acta 56 (2011) 9937–9945
9945
and NiPt3 were still difficult to estimate. The intensities of the
[17] M.J. Valls, J.M. Feliu, A. Aldaz, M.A. Climent, J. Clavilier, J. Electroanal. Chem. 260
1989) 237.
18] J.J. Ruiz, A. Aldaz, M. Dominguez, Can. J. Chem. 56 (1978) 1533.
(
secondary strong peaks in Ni Pt and NiPt3 decreased dramatically
3
[
in Figs. A2–A4. Only the strongest peaks of Ni Pt and NiPt were dis-
3
3
[19] E.L. Pitara, J. Barbier, Appl. Catal. A: Gen. 149 (1997) 49.
cernible in the pattern of the complicated laminate Ni–Pt/Au/Al O .
[20] H.A. Gasteiger, N.M. Markovi c´ , P.N. Ross Jr., J. Phys. Chem. 99 (1995) 16757.
2
3
[
[
[
21] H. Igarashi, T. Fujino, M. Watanabe, J. Electroanal. Chem. 391 (1995) 119.
22] L.D. Burke, L.M. Hurley, Electrochim. Acta 44 (1999) 3451.
23] N.H. Li, S.G. Sun, S.P. Chen, J. Electroanal. Chem. 430 (1997) 57.
The limited information obtained from the finite strongest peaks
could not fit the matrix iteration of the lattice parameters. The exact
solution is to be determined in a subsequent analysis.
[24] X. Xing, I.T. Bae, M. Shao, C.C. Liu, J. Electroanal. Chem. 346 (1993) 309.
[
25] C.E. Dahmani, M.C. Cadeville, J.M. Sanchez, J.L. Morán-López, Phys. Rev. Lett. 55
1985) 1208.
26] J.M. Zen, D.M. Tsai, A.S. Kumar, V. Dharuman, Electrochem. Commun. 2 (2000)
82.
(
References
[
7
[
1] S.J. Padayatty, A. Katz, Y. Wang, P. Eck, O. Kwon, J.H. Lee, S. Chen, C. Corpe, A.
Dutta, S.K. Dutta, M. Levine, J. Am. Coll. Nutr. 22 (2003) 18.
[27] L. Tian, L. Chen, L. Liu, N. Lu, W. Song, H. Xu, Sens. Actuators B: Chem. 113 (2006)
150.
[
[
2] M. Sönmeza, G. Türk, A. Yüce, Theriogenology 63 (2005) 2063.
3] B.A. Eipper, R.E. Mains, C.C. Glembotski, Proc. Natl. Acad. Sci. U.S.A. 80 (1983)
[28] A. Ambrosi, A. Morrin, M.R. Smyth, A.J. Killard, Anal. Chim. Acta 609 (2008) 37.
[29] R. Thangamuthu, S.M.S. Kumar, K.C. Pillai, Sens. Actuators B: Chem. 120 (2007)
745.
5
144.
[
[
4] S.J. Klebanoff, D.D. Dziewiatkowski, G.J. Okinaka, J. Gen. Physiol. 4 (1958) 303.
5] S. Martinoia, G. Massobrio, L. Lorenzelli, Sens. Actuators B: Chem. 105 (2005)
[30] J. Wu, J. Suls, W. Sansen, Electrochem. Commun. 2 (2000) 90.
[31] Y. Sha, L. Qian, Y. Ma, H. Bai, X. Yang, Talanta 70 (2006) 556.
[32] J. Wang, Z. Wang, S. Wang, Synth. Met. 156 (2006) 610.
[33] A.M. Yu, H.Y. Chen, Anal. Chim. Acta 344 (1997) 181.
[34] P.R. Roy, M.S. Saha, T. Okajima, T. Ohsaka, Electrochim. Acta 51 (2006) 4447.
[35] D. Oukil, L. Makhloufi, B. Saidani, Sens. Actuators B: Chem. 123 (2007) 1083.
[36] Q. Sheng, H. Yu, J. Zheng, J. Electroanal. Chem. 606 (2007) 39.
[37] E.S. Ribeiro, L.T. Kubota, Microchim. Acta 154 (2006) 303.
[38] F.A. Pavan, E.S. Ribeiro, Y. Gushikem, Electroanalysis 17 (2005) 625.
[39] R.P. Kalakodimi, M. Nookala, Anal. Chem. 74 (2002) 5531.
[40] Z. Guorong, W. Xiaolei, S. Xingwang, S. Tianling, Talanta 51 (2000) 1019.
[41] H. Okamoto, J. Phase Equilib. Diffus. 31 (2010) 322.
1
4.
[
[
[
6] A. Jain, A. Chaurasia, K.K. Verma, Talanta 42 (1995) 779.
7] S. Ivanow, V. Tsakova, V.M. Mirsky, Electrochem. Commun. 8 (2006) 643.
8] M.K. Amini, S. Hahrokhian, S. Tangestaninejad, V. Mirkhani, Anal. Biochem. 290
(
2001) 277.
[
9] M. Arvand, Sh. Sohrabnezhad, M.F. Mousavi, Anal. Chim. Acta 491 (2003) 193.
[
[
[
[
[
[
[
10] X. Wang, H. Watanabe, S. Uchiyama, Talanta 74 (2008) 1681.
11] J.B. Raoof, R. Ojani, A. Kiani, J. Electroanal. Chem. 515 (2001) 45.
12] G. Moreno, F. Pariente, E. Lorenzo, Anal. Chim. Acta 420 (2000) 29.
13] J.J. Sun, D.M. Zhou, H.Q. Fang, H.Y. Chen, Talanta 45 (1998) 851.
14] M.I. Manzanares, V. Solis, R.H. de Rossi, J. Electroanal. Chem. 430 (1997) 163.
15] M.I. Manzanares, V. Solis, R.H. de Rossi, J. Electroanal. Chem. 407 (1996) 141.
16] A. Serpen, V. Gokmen, Food Chem. 104 (2007) 721.
[42] G. Carturan, G. Cocco, S. Enzo, R. Ganzerla, M. Lenarda, Mater. Lett. 7 (1988) 47.
[43] Y.C. Weng, Y.L. Hsiao, J. Electroanal. Chem. 651 (2011) 160.