scispace - formally typeset

Journal ArticleDOI

Diffusion regimes at nanoelectrode ensembles in different ionic liquids

01 Mar 2010-Electrochimica Acta (Pergamon)-Vol. 55, Iss: 8, pp 2865-2872

Abstract: The electrochemical and diffusion behaviour of different redox probes in different ionic liquids is studied at gold nanoelectrode ensembles (NEEs) in comparison with millimetre sized gold (Au-macro) and glassy carbon (GC) disk electrodes. The redox probes are neutral ferrocene (Fc), the ferrocenylmethyltrimetylammonium cation (FA+) and the ferrocenylmonocarboxylate anion (FcCOO−). The ILs are the dicyanamide, [N(CN)2] or bis(trifluoromethylsulfonyl)amide), [N(Tf)2] salts of the following cations: 1-butyl-3-methylimidazolium, [BMIm], 1-butyl-3-methylpyrrolidonium, [BMPy], or tris(n-hexyl)tetradecylphosphonium [P14,666]. These ILs are characterized by different viscosities, ranging from 32 to 277 cP. The cyclic voltammetric behaviour of the redox probes is reversible and diffusion controlled at GC electrodes. Diffusion coefficients (D) calculated by the Randles–Sevcik equation scales inversely with the IL viscosity, ranging from 2 × 10−8 to 3 × 10−7 cm2 s−1. Ionic solutes, namely FA+ and FcCOO−, present slightly lower D values than neutral Fc. At the Au-macro the electrochemical behaviour of the redox probes is diffusion controlled in the ILs containing the [N(Tf)2] anion, while it involves relevant adsorption processes in the [N(CN)2] containing electrolyte. For this reason the diffusion at gold NEEs is studied only in the former ILs. The CVs of the redox probes at the NEEs are peak shaped at low scan rate ( v ), while they are sigmoidally shaped at high v , but with some shift between forward and backward patterns. This is indicative of the occurrence of a total overlap (TO) diffusion condition when v is low which becomes a mixed diffusion layers (MDL) regime, with only a partial overlapping of individual diffusion layers, at high v values. In the most viscous IL, namely [P14,666] [N(Tf)2], at v higher than 0.8 V s−1, a plateau current independent on the scan rate is achieved, indicating the tendency to reach the pure radial regime in this IL. The v values at which the transition between TO and MDL is observed scales directly with D and inversely with the IL viscosity. This behaviour is interpreted on the basis of the dependence of individual diffusion layers at each nanoelectrode on redox probe/IL interaction which fits with existing theoretical models very recently developed for nanoelectrode arrays.
Topics: Diffusion (52%), Ionic liquid (50%)

Content maybe subject to copyright    Report

Electrochimica Acta 55 (2010) 2865–2872
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Diffusion regimes at nanoelectrode ensembles in different ionic liquids
Paolo Ugo
a,
, Ligia M. Moretto
a
, Manuela De Leo
a
, Andrew P. Doherty
b
,
Chiara Vallese
a
, Sreekanth Pentlavalli
b
a
Department of Physical Chemistry, University of Venice, Santa Marta 2137, 30123, Venice, Italy
b
The School of Chemistry and Chemical Engineering, Queen’s University of Belfast, David Keir Building, Stranmillis Road, Belfast, N.I. BT5AG, UK
article info
Article history:
Received 9 September 2009
Received in revised form
18 December 2009
Accepted 21 December 2009
Available online 11 January 2010
Keywords:
Nanoelectrodes
Arrays
Ionic liquids
Diffusion
Voltammetry
abstract
The electrochemical and diffusion behaviour of different redox probes in different ionic liquids is stud-
ied at gold nanoelectrode ensembles (NEEs) in comparison with millimetre sized gold (Au-macro)
and glassy carbon (GC) disk electrodes. The redox probes are neutral ferrocene (Fc), the ferrocenyl-
methyltrimetylammonium cation (FA
+
) and the ferrocenylmonocarboxylate anion (FcCOO
). The ILs
are the dicyanamide, [N(CN)
2
] or bis(trifluoromethylsulfonyl)amide), [N(Tf)
2
] salts of the following
cations: 1-butyl-3-methylimidazolium, [BMIm], 1-butyl-3-methylpyrrolidonium, [BMPy], or tris(n-
hexyl)tetradecylphosphonium [P
14,666
]. These ILs are characterized by different viscosities, ranging from
32 to 277 cP. The cyclic voltammetric behaviour of the redox probes is reversible and diffusion controlled
at GC electrodes. Diffusion coefficients (D) calculated by the Randles–Sevcik equation scales inversely
with the IL viscosity, ranging from 2 × 10
8
to 3 × 10
7
cm
2
s
1
. Ionic solutes, namely FA
+
and FcCOO
,
present slightly lower D values than neutral Fc. At the Au-macro the electrochemical behaviour of the
redox probes is diffusion controlled in the ILs containing the [N(Tf)
2
] anion, while it involves relevant
adsorption processes in the [N(CN)
2
] containing electrolyte. For this reason the diffusion at gold NEEs is
studied only in the former ILs.
The CVs of the redox probes at the NEEs are peak shaped at low scan rate (
v), while they are sigmoidally
shaped at high
v
, but with some shift between forward and backward patterns. This is indicative of the
occurrence of a total overlap (TO) diffusion condition when
v is low which becomes a mixed diffusion
layers (MDL) regime, with only a partial overlapping of individual diffusion layers, at high
v values. In
the most viscous IL, namely [P
14,666
] [N(Tf)
2
], at v higher than 0.8 V s
1
, a plateau current independent on
the scan rate is achieved, indicating the tendency to reach the pure radial regime in this IL. The
v
values
at which the transition between TO and MDL is observed scales directly with D and inversely with the
IL viscosity. This behaviour is interpreted on the basis of the dependence of individual diffusion layers
at each nanoelectrode on redox probe/IL interaction which fits with existing theoretical models very
recently developed for nanoelectrode arrays.
© 2010 Elsevier Ltd. All rights reserved.
1. Introduction
Novel interfaces in which both the solid and liquid compo-
nents present peculiar characteristics are increasingly used in
electrochemistry. The present study focuses on the electrochemi-
cal behaviour of a quite complex electrode/solution interface where
the electrode is the surface of an ensemble of gold nanoelectrodes
and the electrolyte is an ionic liquid (IL).
ILs are salts composed by a large organic cation and a rel-
atively small anion; they present peculiar properties both from
a fundamental and applicative viewpoint. For the electrochemist
they are intriguing aprotic solvents which can stabilize electro-
Corresponding author.
E-mail address: ugo@unive.it (P. Ugo).
generated reactive species or water insoluble redox probes as
well as allow the wide extension of the accessible potential win-
dow [1–4]. Particularly interesting is the role of ILs in ruling the
mass transport of electroactive analytes to the electrode/solution
interface [5–8]. Moreover, the nature and characteristics of ILs
can influence dramatically the electron transfer kinetics [9,10].
The large differences in size and ionic mobility of the cation and
anion can originate selective solvation of solutes with different
ionic charge. For instance, for the redox couple O
2
/O
2
it was
observed that in some RTIL the diffusion coefficient for the neu-
tral species is significantly larger than the one of the superoxide
anion, this reflecting in a dramatic asymmetry of the voltammet-
ric peak currents for the anodic and cathodic processes [11,12].
In some ILs, a complex concentration dependence of the diffusion
coefficients was observed, at least in certain concentration range
[13,14].
0013-4686/$ see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2009.12.059

2866 P. Ugo et al. / Electrochimica Acta 55 (2010) 2865–2872
Although the electrochemical behaviour of ILs has been quite
widely investigated by using millimetre or micrometre sized elec-
trodes [3,4], only one recent report from our lab [15] introduced
the study of complex nanostructured electrode systems such as
nanoelectrode ensembles (NEEs) in a specific IL, namely, 1-butyl-3-
methylimidazolium tetrafluoroborate, [BMIm][BF
4
]. In particular,
this first investigation indicated that the high viscosity of this ionic
liquid (approximately 100–150 cP) caused a change in diffusion
regime with respect to less viscous electrolyte solution, e.g. aque-
ous or acetonitrile solutions.
The nanoelectrode ensembles used in our laboratory are indeed
prepared by electroless deposition of gold nanodisk electrodes
using commercially available track-etched polycarbonate mem-
branes as templates [16,17]. Templated NEEs prepared by the
procedure introduced by Menon and Martin in 1995 [18], are
increasingly studied and applied by many laboratories [19,20].
NEEs can be considered as partially blocked electrodes
[18,21–23], being composed by a very large number of very small
ultramicroelectrodes, all statistically equivalent since their large
number make border effects negligible [24,25]. From a general
viewpoint, arrays of micro and nanoelectrodes can exhibit dis-
tinct voltammetric response regimes depending on the thickness
of the diffusion layer and distance between the nanoelectrode ele-
ments [16,18,22–25]. Three distinct limit cases can be achieved:
(A) total overlap (TO) regime: when radial diffusion boundary lay-
ers overlap totally (hemi-distance between nanoelectrodes smaller
than radius of diffusion hemispheres); (B) pure radial (PR): when
the nanoelectrodes behave independently (hemi-distance between
nanoelectrodes larger than radius of diffusion hemispheres); (C)
linear active (LA): when the nanoelectrodes behave as isolated
planar electrodes, each under semi-infinite linear diffusion. In prin-
ciple, for a fixed geometry of the array, it could be possible to
move from TO to PR to LA by increasing the voltammetric scan
rate. Intermediate regimes can be also observed: for experimen-
tal conditions between TO and PR the so-called mixed diffusion
layers (MDL) regime can be established. Under these conditions
partial overlap of diffusion layers occurs. Theoretical models have
been developed for ordered [22,25–27] and random [28] arrays of
electrodes.
In water solutions, the diffusion regime usually observed at NEEs
fabricated from commercial track-etched membranes is the TO;
only in the case of custom made polycarbonate membranes, Mar-
tin et al. were able to achieve the PR regime [29]. The limits which
hinder reaching the PR regime with NEEs made with commercially
available track-etch membranes are the following:
- the pores in the membrane (and the templated nanodisk elec-
trodes) are randomly distributed;
- the average distance, d, between the nanoelectrodes is fixed and
ruled by the producer.
This derives from the procedure used to prepare the mem-
branes, that is the tracking of a polycarbonate film with high
energy radiation in a nuclear reactor or ion accelerator, followed
by development of the tracks by etching in alkaline solution [16].
The density of pores per unit surface is determined by the radia-
tion source and the exposure time during the tracking, while the
diameter of the pores is determined by the etching time. Commer-
cially available membranes contain monodisperse pores with 10,
30, 200 nm or larger diameter and with a pore density of 10
8
to
10
9
pores/cm
2
.
At fixed scan rate, the thickness, ı, of the diffusion layer around
each nanoelectrode depends on the square root of the diffusion
coefficient (D) of the redox analyte [30], according to: ı =(˘Dt)
1/2
.
D values are themselves inversely proportional to the viscosity of
the solvent medium, given by the Stokes–Einstein equation [30];as
Table 1
Surveys of the literature data on the viscosities, , of the studied ionic liquids.
IL /cP T/
C Ref.
[BMIm][N(CN)
2
]3228 [32]
[BMPy][N(CN)
2
]5025[33]
[BMIm][N(Tf)
2
]5720[31]
[BMPy][N(Tf)
2
]8925[34]
[P
14,666
][N(Tf)
2
] 277 25 [35]
a consequence, changes in viscosity can reflect in dramatic changes
in the diffusion regime.
In order to investigate with more insight on the role of solvent
viscosity on the diffusion regime at NEEs, in the present work
we study the electrochemical behaviour at NEEs of different
redox analytes in a series of different ionic liquids. The ILs stud-
ied were salts of two different anions, namely, dicyanamide,
[N(CN)
2
], and bis(trifluoromethylsulfonyl)amide), [N(Tf)
2
].
The cations of the salts were 1-butyl-3-methylimidazolium,
[BMIm], 1-butyl-3-methylpyrrolidonium, [BMPy], and tris(n-
hexyl)tetradecylphosphonium [P
14,666
]. As outlined in Table 1,
this combination of anions and cations, furnishes five different ILs
characterized by different viscosities, , which range from 32 to
277 cP.
The analytes are three different redox probes containing the fer-
rocenyl group and differing each other for ionic charge; they are
neutral ferrocene (Fc), ferrocenylmethyltrimetylammonium cation
(FA
+
) and ferrocenylmonocarboxylic acid (FcCOOH). On the basis
of the basicity of the anions [N(Tf)
2
] and [N(CN)
2
]
[36] and of
the acidity of FcCOOH (pK
a
= 6.7 in water/ethanol [37]), it is rea-
sonable that FcCOOH be completely dissociated to the ferrocenyl
carboxylate anion.
Preliminary experiments at millimetre sized electrodes were
performed to measure the diffusion coefficients of the redox probes
in the different ILs as well as to gain information of the general
electrochemical behaviour of the redox probes in the ILs using two
different electrode materials, namely, gold (of which are made the
NEEs) and glassy carbon (GC).
2. Experimental
2.1. Apparatus and procedures
Voltammetric measurements were performed with a CH660A
apparatus controlled via PC, using IR-drop compensation. All elec-
troanalytical measurements were carried out in a three-electrodes
cell of small volume (5 mL). The working electrode was either a NEE,
a glassy carbon electrode (GCE, diameter 5 mm) or an Au-disk elec-
trode, millimetre sized (diameter 3 mm), the counter electrode was
a platinum coil and an Ag wire was used as quasi-reference elec-
trode. ILs were dried overnight in a vacuum oven at 40
C before
use, after treatment with activated molecular sieves. It was pre-
viously demonstrated that this procedure is able to remove trace
water from hygroscopic IL such as [BMIm][BF
4
] [38]. This was
confirmed for the hydrophobic ILs studied here, since no water
reduction peak was observed in the CV recorded after perform-
ing the above described vacuum treatment. All measurements
were performed at room temperature (20 ± 1
C), operating under
a nitrogen atmosphere; the purging gas fluxed trough traps loaded
with concentrated sulphuric acid, in order to prevent eventual
entrance into the cell of humidity from the room environment.
2.2. Preparation of the nanoelectrode ensembles
Polycarbonate filtration membranes (SPI-Pore, 47 mm filter
diameter, 6 m filter thickness) with nominal pore diameter of

P. Ugo et al. / Electrochimica Acta 55 (2010) 2865–2872 2867
Fig. 1. SEM image of a NEE.
30 nm, pore density of 6 × 10
8
pores cm
2
and coated by the pro-
ducer with polyvinylpyrrolidone were used as the templates to
prepare the NEEs, following the Menon and Martin procedure
[18] with recent updates [39]. Briefly, after wetting for 2 h in
methanol, the polycarbonate template membrane was sensitized
with Sn
2+
by immersion into a solution that was 0.026 M in SnCl
2
and 0.07 M in trifluoroacetic acid in 50:50 methanol–water for
5 min. After rinsing with methanol for 5 min, the sensitized mem-
brane was immersed for 10 min in 0.029 M Ag[(NH
3
)
2
]NO
3
. The
membrane was then immersed into the Au plating bath which
was 7.9 × 10
3
MinNa
3
Au(SO
3
)
2
, 0.127 M in Na
2
SO
3
. After wait-
ing 30 min, 0.625 M formaldehyde was added to the plating bath;
this delay time was introduced here since it allows one to sepa-
rate the formation of the first gold nuclei (produced by galvanic
displacement of metallic Ag
0
nuclei with Au
0
nuclei) from the fol-
lowing catalytic growth of these nuclei by further gold deposition
caused by formaldehyde. The temperature of the bath was 0–2
C.
Electroless deposition was allowed to proceed for 15 h, after which
an additional 0.3 M formaldehyde was added. Deposition was con-
tinued for another 9 h, after which the membrane was rinsed with
water and immersed in 10% HNO
3
for 12 h. The membrane was
then rinsed again with water and dried. For the final assembly of
the NEEs, see previously published procedures [16].
The geometric area (area of the nanoelectrodes + area of the
insulator among them) was 0.07 cm
2
, nanodisk diameters mea-
sured by SEM was 40–50 nm [4]. A SEM image of a NEE used in
this work is shown in Fig. 1; for further details see Ref. [40].
Commercial gold electroless plating solution Oromerse Part B
was purchased from Technic Inc.
2.3. Chemicals
(Ferrocenylmethyl)dimethylamine (Aldrich) was reacted with
methyl iodide to form the quaternary ammonium iodide [41]. This
was then converted to (ferrocenylmethyl) trimethylammonium
hexafluorophosphate (FA
+
PF
6
) using AgPF
6
.
Ferrocene (Fc) and ferrocenylmonocarboxylic acid (FcCOOH)
were purchased from Sigma.
[BMIm], [BMPy] and [P
14,666
] ionic liquids were prepare by ion
metathesis reaction of their chloride salts with either NaN(CN)
2
or LiN(Tf)
2
in dichloromethane (DCM) solvent. The procedure
involves firstly dissolving the quaternary ammonium or phospho-
nium cation salt in DCM followed by the addition of 1.2 equivalents
of the anion salt. Reaction mixtures were stirred for 24 h after which
the precipitated inorganic salts (either NaCl or LiCl) were removed
by filtration. The DCM (filtrate) solution of the crude ionic liquid
was washed (5×) with deionised water. After washing, DCM was
removed under vacuum. Activated carbon was then added to the
ionic liquid and stirred for 24 h. Finally, the activated carbon was
removed by filtration through an alumina column under vacuum.
All other reagents were of analytical grade and used as received.
3. Results and discussion
3.1. GC-macro
We examined at first the main features of the electrochemical
behaviour of Fc, FcCOO
and FA
+
in the investigated ILs, focusing
in particular on the reversibility and diffusion control of the redox
process in order to evaluate diffusion coefficient values. This pre-
liminary characterization was performed using millimetre sized GC
disk electrodes. Under these conditions, for all the redox probes
and ILs here examined, a diffusion controlled reversible or quasi-
reversible electrochemical behaviour was observed. As an example,
Fig. 2A shows the typical CV patterns recorded with the GCE at dif-
ferent scan rates in [BMIm][N(CN)
2
] containing 1 mM Fc. For scan
rates 0.050 V/s, the peak to peak separation (E
p
) is 0.065 V and
Fig. 2. (A) Cyclic voltammograms of 1 mM Fc in [BMIm][N(CN)
2
] recorded at GCE
(geometric area 0.19 cm
2
) at different scan rates: 20, 50, 100, 200, 500, 1000 mV s
1
;
peak currents scale with the scan rate. (B) Dependence of the anodic peak current
on the square root of the scan rate.

2868 P. Ugo et al. / Electrochimica Acta 55 (2010) 2865–2872
Table 2
Diffusion coefficients measured by cyclic voltammetry at the glassy carbon electrode in solution containing 1 mM redox probe in different ionic liquids.
D(Fc)/10
7
cm
2
s
1
D(FcCOO
)/10
7
cm
2
s
1
D(FA
+
)/10
7
cm
2
s
1
[BMIm][N(CN)
2
] 3.2 ± 0.3 2.0 ± 0.4 1.2 ± 0.4
[BMPy][N(CN)
2
] 2.5 ± 0.3 2.0 ± 0.4 1.2 ± 0.4
[BMIm][N(Tf)
2
] 2.9 ± 0.5 0.9 ± 0.2 n.m.
a
[BmPy][N(Tf)
2
] 0.7 ± 0.3 0.4 ± 0.2 0.5 ± 0.2
[P
14,666
][N(Tf)
2
] 0.2 ± 0.05 n.m.
a
n.m.
a
a
n.m. = not measured.
E
1/2
, calculated as E
1/2
=(E
pf
+ E
pb
)/2, is independent of the scan rate.
The anodic peak current values (I
pf
) scale linearly with the square
root of the scan rate (see Fig. 2B) and the I
pb
/I
pf
ratio is always very
close to unity (namely, 0.98 ± 0.03).
All these features indicate a diffusion controlled reversible elec-
trochemical behaviour for the oxidation:
Fc Fc
+
+ e (1)
A similar behaviour was observed at the GCE for Fc in the other
ILs as well as for FcCOO
and FA
+
in all the examined solvents. For
the case of the high viscosity IL [P14,666][N(Tf)
2
], only ferrocene
was examined as the electroactive solute. Note that, since an Ag
wire was used as a pseudo-reference electrode, no quantitative
comparison of E
1/2
values of different analytes and/or in different
IL can be performed. From a qualitative viewpoint it was observed
that the oxidation of FcCOO
and FA
+
takes place at potential values
not too far each other and, in almost all the examined ILs, the oxi-
dation of the ionic Fc derivatives occurs at more positive potential
values than the oxidation of neutral Fc.
On the basis of the observed electrochemical behaviour, diffu-
sion coefficients values were calculated by the slopes of the linear
Ip vs.
v
1/2
plots, by applying the Randles–Sevcik equation [29]:
I = 2.69 × 10
5
n
3/2
AD
1/2
Cv
1/2
(2)
Diffusion coefficient values were measured in triplicate and rel-
evant average values ± maximum range are listed in Table 2.
These data indicate that diffusion coefficients values of Fc and
derivatives in the examined ILs are from 1 to 2 orders of magnitude
lower than D values in aqueous [42,43] or acetonitrile solutions
[44]. Comparison between data in Tables 1 and 2 confirms that
D values for the same redox probe decrease significantly with
increasing the solvent viscosity. Because of the variability of the
experimental conditions relevant to the data in Table 1, which come
from different literature sources, care must be taken in trying to
check quantitatively such a dependence.
For D values already available in the literature some consider-
ation can be done. For the case of Fc in [BMIm][N(Tf)
2
] the value
determined by us at 20
C, namely 2.9 × 10
7
cm
2
s
1
, is in reason-
able agreement with the literature values of 3.77 × 10
7
cm
2
s
1
[45], measured at 26
C. Higher temperature means lower viscos-
ity and higher D values; it was reported that each two-degree
shift results in 10% change of the IL viscosity [46], so that, tak-
ing into account this factor, the two values are almost the same.
Note that recently, for the same probe and IL, Vorotyntsev mea-
sured at 20
CaD value of 1.7 × 10
7
cm
2
s
1
[46]. On the other
hand, the D value determined by us for Fc in [BmPy][N(Tf)
2
], namely
0.7 × 10
7
cm
2
s
1
, is about three times smaller than the value
2.31 × 10
7
cm
2
s
1
, measured by Compton’s group [45]. In this
case, taking into account the temperature dependence of D val-
ues reduces the scattering between our and the Compton’s group
datum, but does not cancel it, so that other factors, including
the different measurement technique (that is CV vs. double step
chronoamperometry at ultramicroelectrodes), should be consid-
ered.
In any case, the cross-comparison of our D values, all measured
with the same procedure, indicates that the diffusion coefficients
of the ionic derivatives are smaller than those of neutral ferrocene.
This suggests the occurrence of a stronger interaction of the ILs with
the ionic ferrocenes. In ILs, the lowering of diffusion coefficients
for ionic species with respect to their neutral analogues is indeed
documented in the literature [11,12]. Compton et al. demonstrated
[45], that the D value for neutral ferrocene is much larger than
D for the ferricinium cation, the same holding also for the cobal-
tocene/cobalticinium couple. A comparable unequality in diffusion
coefficients of the butylviologen dication vs. the mono-cationic rad-
ical was reported to be the cause of differences in reduction peak
currents for the two species in [BMIm][BF
4
] [15].
3.2. Au-macro
The electrochemical behaviour of neutral Fc, FcCOO
and FA
+
at the Au-macro is reversible or quasi-reversible in the ILs con-
taining the [N(Tf)
2
]
anion, with features comparable with those
obtained at the GC-macro. On the other hand, in the ILs contain-
ing the [N(CN)
2
]
anion significant differences and complications
are observed at the Au-macro. Fig. 3 shows the CVs for FcCOO
in
[BMIm][N(CN)
2
].
An oxidation peak with almost symmetrical shape is observed
with the peak potential shifting from 0.660 to 0.710 V when increas-
ing the scan rate. In the backward scan a small reduction peak
is observed at about 0.570 V. The oxidation peak scales linearly
with the scan rate (see insert in Fig. 3), in agreement with the
occurrence of an adsorption controlled process [29]. This can
be explained taking into account that the [N(CN)
2
]
anion can
adsorb on gold, modifying the electrode surface and acting as
a bridge in the following adsorption and oxidation of the redox
probe.
The situation observed for ferrocene in the same IL is shown in
Fig. 4. Now, in the oxidation scan two peaks are observed, sug-
gesting the occurrence of two following processes. The current
of the first anodic peak depends linearly on the square root of
the scan rate, while the current of the second anodic peak scales
directly with
v (not shown). As a consequence of such a different
dependence on the scan rate, the first peak is the prevailing one
at low scan rates (e.g. 0.020 V s
1
) while the second peak increases
when increasing the scan rate, up to being very well detected at
v > 0.5Vs
1
. These evidences suggest that the first peak is related
to the oxidation of diffusing Fc while second peak can be due to
adsorption on Au, via adsorbed [N(CN)
2
]
, of electrogenerated Fc
+
.
Similarly complex voltammetric patterns were observed at the Au-
macro for the other redox probes in [BMPy][N(CN)
2
].
In the following part of this research we were interested in
investigating mainly the role of the solvent on the diffusion of the
analyte to the nanoelectrodes, therefore we chose to avoid any
complication by adsorption or other non-diffusion controlled pro-
cesses by focusing our attention on the behaviour of ferrocene and
its derivatives in the [N(Tf)
2
] containing ILs.
3.3. Voltammetry with NEEs
Fig. 5A shows the CVs recorded at three different scan rates with
a NEE in 5 mM Fc in [BMPy][N(Tf)
2
].

P. Ugo et al. / Electrochimica Acta 55 (2010) 2865–2872 2869
Fig. 3. (A) Cyclic voltammograms of 1 mM FcCOOH in [BMIm][N(CN)
2
] recorded at
an Au-macro electrode (geometric area 0.07 cm
2
), at different scan rates: 10, 20,
50, 100, 200, 400, 800 and 1000 mV s
1
; peak currents scale with the scan rate. (B)
Dependence of the anodic peak current on the scan rate.
Fig. 4. Cyclic voltammograms of 1 mM Fc in [BMIm][N(CN)
2
] recorded at an
Au-macro electrode at different scan rates: 50, 100, 200 and 500 mV s
1
; other
parameters as in Fig. 3.
At low scan rates peak shaped voltammograms are recorded
which become sigmoidally shaped at higher scan rates, however
with a shift detectable between the forward and backward pat-
terns. As shown in Fig. 5B and C, similar behaviours are observed
in [BMIM][N(Tf)
2
] and [P
14,666
][N(Tf)
2
], respectively; note also that
similarly sigmoidally shaped CV patterns were previously observed
at NEEs for the case of FA
+
or heptylviologen cations in [BMIm][BF
4
]
[15]. On the basis of recent reports [26], the CV shape observed at
high scan rates could be considered typical of the mixed diffusion
layers regime, which is observed when there is a partial overlapping
of individual diffusion layers. However, even the contribution of the
double-layer charging current (I
c
) cannot be neglected, since such
a capacitive current can also be responsible for the lack of overlap
between the forward and backward voltammetric patterns. Note
that in IL double-layer charging currents are significantly higher
than that in water. Moreover, to reach the pure radial regime with
NEEs it is necessary to operate at relatively high scan rates and I
c
scales directly with v; for instance, for our NEEs at 500 mV s
1
, I
c
is
estimated around 300 nA in [P
14,666
][N(Tf)
2
] vs. 60 nA in water.
In Fig. 5A some sloping of the CV pattern is also observed. This
could be due to some slowness of the electron transfer kinetics,
but also the contribution of some residual un-compensated IR-drop
cannot be excluded; very high viscosity IL such as [P
14,666
][N(Tf)
2
]
are indeed characterized also by very high ohmic resistance [47]
which cannot necessarily be fully corrected by instrumental IR-
drop compensation.
Plots in Fig. 6 report the typical dependence of the maximum
current (that is the plateau current or the peak current at high or low
scan rates, respectively) as a function of the scan rate (A) and square
root of the scan rate (B). In order to make the comparison clearer,
currents were normalized over the maximum current recorded for
each set of measurements in the same IL.
At very low scan rates, the peak current increases linearly with
v
1/2
, indicating the operativity of total overlap conditions. At very
high scan rates, the current becomes practically independent of the
scan rate, while at intermediate scan rate a complex dependence
is operative. These plots put in evidence that the maximum value
of the scan rate up to which I
p
increases linearly with v
1/2
depends
on the nature of the IL, scaling inversely with the solvent viscosity.
Such a
v
maximum is approximately 50 mV s
1
in [P
14,666
][N(Tf)
2
]
and 90 mV s
1
in [BMPy][N(Tf)
2
]; data in [BMIm][N(Tf)
2
] are more
scattered.
The plots show that also the scan rate at which the current
responses start to be independent of the scan rate, depends on the
nature of the ILs. Note that the observation of such an independency
is indicative of a strong tendency to the PR regime, particularly
evident for [P
14,666
][N(Tf)
2
]. Again, the scan rate at which such a sit-
uation is achieved scales inversely with the solvent viscosity, that is
higher is the viscosity, lower is the scan rate at which independency
starts to be detected.
A very practical graphical representation of the transitions
between different diffusion regimes at ordered arrays of micro-
electrodes was introduced very recently by Guo and Lindner [26].
Even if NEEs are random distributed, some interesting considera-
tion can be extrapolated (at least semi-quantitatively) also for the
NEE case. The limit of the different diffusion zones is defined on
the basis of values of two dimensionless parameters, namely, the
dimensionless distance:
d
=
d
a
(3)
where d is the center-to-center distance between adjacent
microelectrodes, and a is the microelectrode radius; and the dimen-
sionless scan rate:
V =
nF
va
2
(4RTD)
(4)

Citations
More filters

Journal ArticleDOI
TL;DR: The design of a novel immunosensor and its application for celiac disease diagnosis, based on an electrogenerated chemiluminescence (ECL) readout, using membrane-templated gold nanoelectrode ensembles (NEEs) as a detection platform, showing to be suitable to discriminate between healthy and celiac patients.
Abstract: We report here the design of a novel immunosensor and its application for celiac disease diagnosis, based on an electrogenerated chemiluminescence (ECL) readout, using membrane-templated gold nanoelectrode ensembles (NEEs) as a detection platform. An original sensing strategy is presented by segregating spatially the initial electrochemical reaction and the location of the immobilized biomolecules where ECL is finally emitted. The recognition scaffold is the following: tissue transglutaminase (tTG) is immobilized as a capturing agent on the polycarbonate (PC) surface of the track-etched templating membrane. It captures the target tissue transglutaminase antibody (anti-tTG), and finally allows the immobilization of a streptavidin-modified ruthenium-based ECL label via reaction with a suitable biotinylated secondary antibody. The application of an oxidizing potential in a tri-n-propylamine (TPrA) solution generates an intense and sharp ECL signal, suitable for analytical purposes. Voltammetric and ECL analy...

53 citations


Journal ArticleDOI
TL;DR: This review deals with recent advances in bioelectroanalytical applications of nanostructured electrodes, in particular nanoelectrode ensembles (NEEs) and arrays (NEAs), and nanofabrication techniques, principles of function, and specific advantages and limits of NEEs and NEAs.
Abstract: This review deals with recent advances in bioelectroanalytical applications of nanostructured electrodes, in particular nanoelectrode ensembles (NEEs) and arrays (NEAs). First, nanofabrication techniques, principles of function, and specific advantages and limits of NEEs and NEAs are critically discussed. In the second part, some recent examples of bioelectroanalytical applications are presented. These include use of nanoelectrode arrays and/or ensembles for direct electrochemical analysis of pharmacologically active organic compounds or redox proteins, and the development of functionalized nanoelectrode systems and their use as catalytic or affinity electrochemical biosensors.

43 citations


Cites background from "Diffusion regimes at nanoelectrode ..."

  • ...[67]) Bioelectroanalysis with nanoelectrode arrays 3721...

    [...]

  • ...It has recently been shown that, for NEE, transition from the total overlap to the pure radial diffusion can be observed on increasing electrolyte viscosity [67]....

    [...]


Journal ArticleDOI
06 May 2011-Nanotechnology
TL;DR: Electrochemical results showed satisfactory agreement between experimental voltammograms and suitable theoretical models, and the peculiarities of NEAs versus ensembles of nanoelectrodes, obtained by membrane template synthesis, are critically evaluated.
Abstract: Ordered arrays of nanoelectrodes for electrochemical use are prepared by electron beam lithography (EBL) using polycarbonate as a novel e-beam resist. The nanoelectrodes are fabricated by patterning arrays of holes in a thin film of polycarbonate spin-coated on a gold layer on Si/Si3N4 substrate. Experimental parameters for the successful use of polycarbonate as high resolution EBL resist are optimized. The holes can be filled partially or completely by electrochemical deposition of gold. This enables the preparation of arrays of nanoelectrodes with different recession degree and geometrical characteristics. The polycarbonate is kept on-site and used as the insulator that separates the nanoelectrodes. The obtained nanoelectrode arrays (NEAs) exhibit steady state current controlled by pure radial diffusion in cyclic voltammetry for scan rates up to approximately 50 mV s − 1. Electrochemical results showed satisfactory agreement between experimental voltammograms and suitable theoretical models. Finally, the peculiarities of NEAs versus ensembles of nanoelectrodes, obtained by membrane template synthesis, are critically evaluated.

38 citations


Journal ArticleDOI
Pei-Hua Li1, Pei-Hua Li2, Yi-Xiang Li2, Yi-Xiang Li1  +12 moreInstitutions (2)
Abstract: Herein, combined the synergetic effects and excellent adsorption of binary oxides, a novel sensor interface was designed by using porous Ce-Zr oxide nanospheres modified glassy carbon electrode (Ce-Zr oxide/GCE), which has successfully realized high sensitivity and anti-interference detection of Pb(II). The electrochemical determination of Pb(II) has been investigated by square wave anodic stripping voltammetry (SWASV) range from 0.02 to 0.5 μM, and a high sensitivity per unit area of 1666.02 μA μM−1 cm−2 was obtained with a limit of detection of 0.006 μM (3σ method). The enhancement of Pb(II) stripping signal is attributed to the excellent adsorption performance of porous Ce-Zr oxide nanospheres, which has been confirmed with X-ray photoelectron spectroscopy (XPS). Importantly, Ce-Zr oxide/GCE possesses highly anti-interference ability against the influence of Hg(II), Cd(II), Cu(II), and Zn(II) in the determination of Pb(II). Meanwhile, the remarkable stability and reproducibility were obtained. Finally, the accurate analysis of Pb(II) in wastewater collected from Wangtang sewage disposal plant was achieved. These results indicate that porous Ce-Zr oxide nanospheres are identified as promising modifier for the reliable and accurate determination of Pb(II).

32 citations


Journal ArticleDOI
Meng Yang1, Meng Yang2, Xing Chen1, Jinhuai Liu1  +2 moreInstitutions (2)
Abstract: Determination of arsenite [As(III)] without interference in mild condition is crucial for portably assessing arsenic contamination using electrochemical method. We have developed high-surface area nanoporous gold (np-Au) with a three dimensional, interconnected ligaments and nanoporous structure for the electrochemical detection of As(III) in 0.1 M HAc-NaAc solution (pH 5.0) without using strong acidic electrolyte. Square wave anodic stripping voltammetry (SWASV) using the np-Au modified glassy carbon electrode (GCE) confirms the successful detection of As(III) with almost no interference from some commonly coexisting ions. Furthermore, the sensitivity of the np-Au modified GCE exhibited approximately 10-fold enhancement as compared to Au nanoparticles (Au NPs) modified GCE. Finally, the proposed method is successfully applicable for analysis of As(III) in real water samples with satisfactory recoveries. The np-Au modified GCE shows enhanced anti-interference and excellent sensing performance may be attribute to its special surface structure and the fast transports of analytes and electron in the interface of electrode.

30 citations


References
More filters

Journal ArticleDOI
TL;DR: New, hydrophobic ionic liquids with low melting points (<−30 °C to ambient temperature) have been synthesized and investigated, based on 1,3-dialkyl imidazolium cations and hydrophilic anions and thus water-soluble.
Abstract: New, hydrophobic ionic liquids with low melting points (<−30 °C to ambient temperature) have been synthesized and investigated, based on 1,3-dialkyl imidazolium cations and hydrophobic anions. Other imidazolium molten salts with hydrophilic anions and thus water-soluble are also described. The molten salts were characterized by NMR and elemental analysis. Their density, melting point, viscosity, conductivity, refractive index, electrochemical window, thermal stability, and miscibility with water and organic solvents were determined. The influence of the alkyl substituents in 1, 2, 3, and 4(5)-positions on these properties was scrutinized. Viscosities as low as 35 cP (for 1-ethyl-3-methylimidazolium bis((trifluoromethyl)sulfonyl)amide (bis(triflyl)amide) and trifluoroacetate) and conductivities as high as 9.6 mS/cm were obtained. Photophysical probe studies were carried out to establish more precisely the solvent properties of 1-ethyl-3-methylimidazolium bis((trifluoromethyl)sulfonyl)amide). The hydrophobi...

3,951 citations


Journal ArticleDOI
Abstract: Salts having a low melting point are liquid at room temperature, or even below, and form a new class of liquids usually called room temperature ionic liquids (RTIL). Information about RTILs can be found in the literature with such key words as: room temperature molten salt, low-temperature molten salt, ambient-temperature molten salt, liquid organic salt or simply ionic liquid. Their physicochemical properties are the same as high temperature ionic liquids, but the practical aspects of their maintenance or handling are different enough to merit a distinction. The class of ionic liquids, based on tetraalkylammonium cation and chloroaluminate anion, has been extensively studied since late 1970s of the XX century, following the works of Osteryoung. Systematic research on the application of chloroaluminate ionic liquids as solvents was performed in 1980s. However, ionic liquids based on aluminium halides are moisture sensitive. During the last decade an increasing number of new ionic liquids have been prepared and used as solvents. The general aim of this paper was to review the physical and chemical properties of RTILs from the point of view of their possible application as electrolytes in electrochemical processes and devices. The following points are discussed: melting and freezing, conductivity, viscosity, temperature dependence of conductivity, transport and transference numbers, electrochemical stability, possible application in aluminium electroplating, lithium batteries and in electrochemical capacitors.

2,061 citations


Journal ArticleDOI
Philippe Hapiot1, Corinne LagrostInstitutions (1)
20 Jun 2008-Chemical Reviews
TL;DR: A review.
Abstract: A review. Physicochem. properties of ionic liqs. are discussed. Chem. and electrochem. reactivity in ionic liqs. is described including electrode reactions, electrode reaction kinetics, electrosynthesis, etc.

993 citations


Journal ArticleDOI
Abstract: Non-linear diffusion effects induced by partial blocking of an electrode surface toward electron transfer are analyzed in the case where the blocking film is sprinkled with a large number of microscopic, active sites. Under such conditions, the average size of the active sites and the average distance between them are small compared to the total diffusion layer. As a result, non-linear diffusion is confined to a layer adjacent to the electrode surface which is thin compared to the total diffusion layer. It follows that, in the framework of relaxation electrochemical techniques, non-linear diffusion can be regarded as occurring under stationary conditions while linear diffusion is time-dependent. The resolution of the diffusion problem is thus greatly simplified, leading to the description of the polarization as a function of only two dimensionless parameters which are simply related to the experimental parameters: fractional coverage, distance between active sites, standard rate constant of electron transfer and time-range. The formulations thus obtained give evidence for a formal similarity with the polarization problem corresponding to the coupling of electron transfer with preceding and following chemical reactions. According to the values of these parameters, two main types of behavior are obtained: in one of them, the polarization curves show a quasi-reversible behavior, while in the other, kinetic-type waves are obtained. The effect of the various parameters on the characteristics of these waves and on the passage from one type of behavior to the other is described. Procedures for deriving estimates of the size of the active sites and the distance between them from the polarization curves are proposed.

712 citations


Journal ArticleDOI
Abstract: : An electroless deposition procedure for filling the pores in nanoporous filtration membranes with metal (gold) nanowires is described. This method allows us to routinely prepare ensembles of gold nanodisk electrodes in which the nanodisks have diameters as small as 10 nm. Results of electrochemical experiments at ensembles of 30 nm-diameter and 10 nm-diameter gold-disk electrodes are described. The electrochemical response characteristics of these nanoelectrode ensembles are in agreement with predictions of the relevant electrochemical theories. Cyclic voltammetric detection limits for electroactive species at ensembles containing 10 nm-diameter gold disks can be as much as 3 orders of magnitude lower than at a large-diameter gold-disk electrode.

685 citations


Performance
Metrics
No. of citations received by the Paper in previous years
YearCitations
20212
20201
20194
20182
20173
20162