Functional Electro-materials Based on Ferricyanide Redox-active Ionic Liquids

Article abstract of DOI:10.1016/j.electacta.2017.05.201

The unique physical and chemical properties of conventional room temperature ionic liquids (RTILs) render them highly deployable materials in electrochemical devices performing functions such as solvent-free electrolytes in capacitors, batteries and sensors. However, these non-faradaic applications can be complimented by incorporating faradaic redox functionality into the ionic liquid structure which facilitates access to a large array of new electrochemical applications such as dye sensitised solar cells, redox batteries, hydrid capacitors and selective amperometric sensor applications which are all reliant on heterogeneous or homogenous electron-transfer processes. This paper presents and discuses some examples of redox active ionic liquids base on the ferri-/ferro- functionality. These functional electro-materials which are already known [Ref. [18]] exhibit simple reversible one-electron electrochemistry at very negative potentials (by at least ?1?V relative to aqueous systems) in anhydrous media. Glass transition temperatures lower than ?50?°C were also observed along with an overall thermal stability up to at least 400?°C under dry N₂ atmosphere conditions. Opportunities and challenges for these types of electro-materials are discussed.

Full text of DOI:10.1016/j.electacta.2017.05.201

Accepted Manuscript
Title: Functional Electro-materials Based on Ferricyanide
Redox-active Ionic Liquids
Authors: Andrew P. Doherty, Louise Graham, Klaudia
Wagner, David L. Officer, Jun Chen, Gordon G. Wallace
PII:
S0013-4686(17)31217-3
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.electacta.2017.05.201
EA 29647
To appear in:
Electrochimica Acta
Received date:
Revised date:
Accepted date:
16-12-2016
28-5-2017
28-5-2017
Please cite this article as: Andrew P.Doherty, Louise Graham, Klaudia
Wagner, David L.Officer, Jun Chen, Gordon G.Wallace, Functional Electro-
materials Based on Ferricyanide Redox-active Ionic Liquids, Electrochimica
Actahttp://dx.doi.org/10.1016/j.electacta.2017.05.201
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Functional Electro-materials Based on Ferricyanide Redox-active Ionic Liquids
Andrew P. Dohertya,*, Louise Graham a,, Klaudia Wagnerb, David L. Officerb,
Jun Chenb and Gordon G. Wallaceb
a, * School of Chemistry and Chemical Engineering, The Queens’s University of
Belfast, Belfast, Northern Ireland, BT9 5AG, United Kingdom.
e-mail a.p.doherty@qub.ac.uk, telephone + 44 (0)2890974481, fax + 44
(0)28906524
b Intelligent Polymer Research Unit, Australian Research Council Centre of
Excellence for Electromaterials Science, University of Wollongong, Wollongong,
NSW, Australia
1

Abstract
The unique physical and chemical properties of conventional room
temperature ionic liquids (RTILs) render them highly deployable materials in
electrochemical devices performing functions such as solvent-free electrolytes in
capacitors, batteries and sensors. However, these non-faradaic applications can be
complimented by incorporating faradaic redox functionality into the ionic liquid
structure which facilitates access to a large array of new electrochemical applications
such as dye sensitised solar cells, redox batteries, hydrid capacitors and selective
amperometric sensor applications which are all reliant on heterogeneous or
homogenous electron-transfer processes. This paper presents and discuses some
examples of redox active ionic liquids base on the ferri-/ferro- functionality. These
functional electro-materials which are already known [Ref. 18] exhibit simple
reversible one-electron electrochemistry at very negative potentials (by at least -1 V
relative to aqueous systems) in anhydrous media. Glass transition temperatures
o
lower than -50 C were also observed along with an overall thermal stability up to at
o
least 400 C under dry N2 atmosphere conditions. Opportunities and challenges for
these types of electro-materials are discussed.
Key words: electro-materials; task specific ionic liquids; redox active; applications;
electrochemical devices
2

1. Introduction
The acetate ([OAc]) and tetrafluroborate ([BF4]) salts depicted in Figure 1 are
examples of the first reported air and moisture stable RTILs [1].
FIGURE 1 HERE
The introduction of these materials represented a technical step-change from the
earlier unstable chloroaluminate-quaternary ammonium based molten salts [2, 3] and
thus opened up a vast array of chemical and electrochemical possibilities which have
been exploited ever since [4-6]. The key physical or physiochemical properties of
RTILs which make them attractive for electrochemical applications include:
i)
high charge carrier density (>> 1 mol L-1),
inherent ionic conductivity (typically mS cm-1 range),
chemically inertness / extreme redox robustness,
non-volatile / non-flammable,
ii)
iii)
iv)
v)
vi)
thermal stability,
liquid state at low temperatures.
The early work on RTILs concentrated on their use as “green” reaction media [4,
6] but it soon became clear that RTILs could perform useful functions (or “tasks”)
such as directing the course of chemical reactions [7]. So, in addition to the
favourable properties outlined above, the idea of deliberately incorporating chemical
functionality into the RTIL structure to create task-specific ionic liquids (TSILs)
emerged [8, 9] which has contributed greatly to the development of unique
applications beyond the simple electrolytes or solvents paradigms.
It is worth noting that prior to Davis’ 2004 paper detailing Hg2+ sequestering
TSILs, [8], examples of RTILs with built-in electro-active functionality were already
3

known. For example, Murray et al. incorporated both quinone [10] and metal
complexes [11] into “molten salt” structures while viologen IL materials were reported
in 2003 [12]. Contemporaneous to Davis [8], works by Shreeve et al. [13] and Park et
al. [14] involving the incorporating of ferrocene (Fc) moieties (as functionalised
cations) and I- (as anions), respectively, were reported. Subsequently, Wu et al. [15]
reported the incorporation of the redox catalyst TEMPO (2,2,6,6-
tetramethylpiperidine-1-oxyl) into imidazole-based salts which, although solid in their
native form at RT, were soluble in conventional ILs. More recently, Strehmel et al.
[16] reported another TEMPO-based RTIL which acted as a radical probe in ILs.
It is logical that these early examples emerged from classical molecular
electrochemistry and that they point in the direction of important technical
-
applications such as enzyme redox catalysis (Fc/Fc+), dye sensitised solar cells (I-/I3
) (DSSCs), hydrogen peroxide generation / O2 sensing (quinone/hydroquinone) and
alcohol oxidations (TEMPO/nitrosonium). Presenting these, and other well-known
electrochemistries, in appropriate TSIL form may lead to new or improved
electrochemical technologies.
A physical feature of RTILs is their relatively large viscosities (g cm-1 s-1)
compared to molecular solvents. Viscosities typically range from ca. 20 cP to 1000s
of cP with the common imidazolium and pyrrolidinium ILs (depending on the anion)
exhibit values in the moderate 20 – 600 cP range. Deleterious effects of high
viscosity include slow diffusional mass transfer rates as expressed through the
Stokes-Einstein relation, and low ionic conductivities. For the latter, Gratzel et al. [17]
have shown that ion conductivity (S cm-1) of RTILs is an inverse function of the
liquids’ bulk and micro-viscosities. Since all electrochemical devices rely on mass
transport in one way or another minimising viscosity is critical.
4

Since viscosity is intrinsically dependent on the complexity (size, structure and
structural symmetry etc.) of the cation, and the nature of the anion (e.g. H-bonding
ability), the design of new functional TSILs requires appropriate careful
consideration. From an intuitive empirical perspective low viscosity is achieved with
small non-coordinating anions paired with larger, but not too large, structurally
asymmetric cations where charge delocalisation is extensive for both ions. Even for
simple RTILs, they only exist as useful liquids over a fairly narrow range of structures
[19]. Obviously, adding functionality to RTILs inevitably increases mass, and
frequently increases structural symmetry and inter-molecular attractive forces, all of
which favour solidification; these restrictions impose severe challenges to the
molecular design of TSILs.
In addition to minimising viscosity the following functional requirements must
also be met for redox-active ionic liquids:
•
bear a permanent charge i.e. be a RTIL irrespective of the redox state
of the electro-active functionality,
•
•
•
exhibit reversible electrochemistry with an Eo appropriate for the task,
be otherwise inert,
be either hydrophobic or hydrophilic as the application requires.
Although there are many well-known redox chemistries to choose from the
syntheses, characterisation and electrochemistry of ferricyanide ([Fe(CN)6]3-) and
ferrocyanide ([Fe(CN)6]4-) ILs (Figure 2) will be presented here, which follows on
from earlier work [18]. Although ostensibly a simple one-electron redox couple it
exhibits strong dependence on solvation and H-bonding therefore it may act as a
marker for the physiochemical conditions in RTIL environments as well as being of
practical importance in real electrochemical devices and processes.
5

FIGURE 2 HERE
2. Experimental
2.1 Synthesis of trihexyl(tetradecyl)phosphonium bis(trifluoromethanesulfonyl)imide,
[P14, 6,6,6][NTf2] ionic liquid.
In
a
500 cm3 round bottomed flask 69.04
g
(0.132 mol) of
trihexyl(tetradecyl)phosphonium chloride ([P14,,6,6,6][Cl]), Cytec, USA) and 38.04 g
(0.132 mol) of lithium bis(trifluoromethanesulfonyl)imide (3M) were mixed along with
170 cm3 dichloromethane (DCM) and left to stir over night at room temperature. The
LiCl precipitate was then removed by filtration (No. 4 sintered glass filter) and the
resultant filtrated was washed x 6 with deionised water (DI, 18 m cm-1). The DCM /
[P14, 6,6,6][NTf2] solution was then dried with anhydrous MgSO4 and filtered. Finally,
the DCM was removed by evaporation to produce the colourless and highly
hydrophobic [P14,6,6,6][NTf2] ionic liquid.
2.2 Synthesis of trihexyl(tetradecyl)phosphonium ferricyanide, [P14,6,6,6]3[Fe(CN)6]3-
and trihexyl(tetradecyl)phosphonium ferrocyanide, [P14,6,6,6]4[Fe(CN)6]4-
For [P14,6,6,6]3[Fe(CN)6]3- equal volumetric quantities (10 cm3) of acetonitrile (ACN ,
Sigma-Aldrich) and DI water were mixed together thoroughly (resultant volume of
17.8 cm3) and allowed to equilibrate to room temperature after which 11.0 g (0.021
mol) of [P14,,6,6,6][Cl] were added with continued stirring to form a single isotropic
phase. Next, 2.32 g (0.007 mol) of K3[Fe(CN)6] (Sigma-Aldrich) were added to the
ACN/H2O/[P14,6,6,6][Cl] mixture whereupon an immediate ion metathesis occurred
(Equation 1) as evident by a spontaneous “salting-out” driven phase separation
6

process that occurred forming an ACN / [P14,6,6,6]3[Fe(CN)6]3- upper organic phase
and an aqueous KCl (dissolved) bottom phase.
K3[Fe(CN)6] + 3 [P14,6,6,6][Cl]
[P14,6,6,6]3[Fe(CN)6] + 3 KCl (aq) (1)
The organic layer was recovered using a separating funnel and the ACN removed by
evaporation to produce the highly hydrophobic [P14,6,6,6]3[Fe(CN)6]3- ionic liquid which
was subsequently washed x 6 with DI water. Various ACN:H2O ratios were
investigated as were the effects of replacing ACN with acetone and dimethyl
formamide (DMF), neither of which are reported here except for the viscosities
reported in Figure 3.
For the [P14,6,6,6]4[Fe(CN)6]4- synthesis, a slightly different procedure was adopted
due to the limited solubility of K4[Fe(CN)6].3H2O in a 50:50 ACN/H20 mixture. 2.97 g
(0.0070 mol) of K4[Fe(CN)6].3H2O was dissolved in 10 cm3 of DI water. 14.62 g
(0.028 mol) of [P14,6,6,6][Cl] was dissolved in ACN. The aqueous solution was then
slowly added to the ACN solution with stirring whereupon the ion metathesis reaction
shown in Equation 2 occurred resulting in a spontaneous bi-phasic separation.
K4[Fe(CN)6].3H2O + 4[P14,6,6,6][Cl]
where x + y = 4.
[P14,6,6,6]4[Fe(CN)6].3H2O + xKCl (aq) + yKCl(s) (2)
The upper organic phase contained ACN / [P14,6,6,6]4[Fe(CN)6]4- while the bottom
aqueous phase was a saturated KCl solution in contact with copious quantities of
KCl precipitate. The ACN layer was recovered using a separating funnel after which
ACN was removed by evaporation to produce the highly hydrophobic
7

[P14,6,6,6]4[Fe(CN)6]4- ionic liquid which was subsequently washed x 6 with DI water
and dried with anhydrous MgSO4.
2.3 Differential Scanning Calorimetry (DSC) and Thermo-gravimetric analysis (TGA)
DSC analysis was performed using a Perkin Elmer Diamond™ DSC instrument
o
o
over the temperature range -100 oC to 50 C at temperature ramp rate of 100 C /
min. TGA analysis was carried out using a Mettler Toledo TGA/DSC1 Stare
instrument. The sample was heated from room temperature to 800 oC at 10 oC / min.
For both techniques, the samples were held under N2 atmosphere during
measurement.
2.4 Rheological Analysis
o
The viscosity of [P14,6,6,6]4[Fe(CN)6]4- has been reported to be 5790 cP at 20 C
[18]. Viscosity measurements on [P14,6,6,6]3[Fe(CN)6]3- were made using a Brookfield
DV-I digital viscometer fitted with a CP-52 spindle/cup that was attached to a
recirculating water bath (Grant Instruments Ltd) to provide temperature control. The
[P14,6,6,6]3[Fe(CN)6]3- sample was further dried on a Schlenk line under high vacuum
o
at 90 C for 16 hrs prior to measurement which were carried out over the
o
temperature range 25  0.1 C to 80  0.1 oC at 5 oC increments. Care was taken to
ensure the sample had reached equilibrium temperature before a reading was taken
(ca. 20 mins). It was observed that the viscosity was independent of spindle rotation
speed therefor readings at several different speeds at each temperature were taken
and an average obtained.
2.1 Electrochemical Measurements.
8

All electrochemical measurements were performed using a Digi Ivy 2100 mini-
potentiostat in the 3-electrode configuration. The working electrode was a 3 mm
diameter glassy carbon disk (from Bioanalytical Systems (BAS)) which was polished
with 0.05 m -alumina as an aqueous slurry prior to use. The counter electrode was
Pt wire (Goodfellow) sealed in glass. A Pt wire also acted as a quasi-reference
reference electrode with added ferrocene acting as an internal potential reference.
Details of electrolyte composition are given in the text and/or Figure legends as
appropriate.
Electrochemical
characterisation
was
carried
out
on
[P14,6,6,6]3[Fe(CN)6]3- and [P14, 6,6,6]4[Fe(CN)6]4- in both acetonitrile solvent (with 0.1
mol L-1 tetrabutyl ammonium tetrafluoroborate (TBATFB) electrolyte) and in [P14,
o
6,6,6][NTf2] ionic liquid at room temperature (ca. 20 C.) All solutions were sparged
with dry N2 for at least 20 minutes prior to measurement and kept under a constant
N2 blanket during measurement.
For electrochemical measurements under aqueous condition the surfactant
cetyltrimethylammonium bromide (CTAB) was used to solubilise the Fe(CN)6]3-/4- ILs
in an aqueous electrolyte solution that was composed of 0.1 M KCl, 5.0 % (w/v)
CTAB and 0.1 mol L-1 phosphate buffer at pH 7.4. In order to solubilise the
[P6,6,6,14]3[Fe(CN)6]3-, and [P6,6,6,14]4[Fe(CN)6]4-, i.e. to form clear isotropic solutions, it
was necessary to sonicated the mixtures for up to 30 minutes.
Due to uncompensated solution resistance, and the possibility of sluggish
heterogeneous electron transfer kinetics, diffusion coefficients obtained from cyclic
voltammetric data should be considered as indicative only.
3. Results and discussion
3.1 Viscosity measurements
9

Because dynamic events in liquids phases are frequently controlled by the rate of
viscous flow processes the rheological properties of liquids materials is of the utmost
importance – this is of particular consideration in electrochemical devices where
ionic mobility, solvent reorganisation, and diffusional fluxes control the performance
of the devices. The viscous behaviour of RTILs has been the subject of intense
interest and scrutiny over many years [19, 20]. Seddon et al. [19] have shown that
the presence of small quantities of imputes can have a dramatic effect on viscosity.
Gratzel et al. [17] established the relationship between ionic conductivities and
viscosities in RTILs, while Paduszyński and Domańska [20] reported the use of
group contribution models for predicting RTIL viscosity. Given the large molecular
mass and structural symmetry of the ferri- and ferro-cyanide ILs studied here, highly
viscous behaviour is expected and, indeed, is found.
As mentioned earlier, the viscosity of [P14,6,6,6]4[Fe(CN)6]4- has already been
o
reported as 5790 cP at 20 C [18]; we therefore restrict our discussion here to the
viscous behaviour of [P14,6,6,6]3[Fe(CN)6]3-. Figure 3 shows Arrhenius plots for the
viscosity of this IL prepared by four different methods, the line is for eye guidance
only and is not a product of linear regression.
Figure 3 here
It is immediately evident from the data that the ILs’ viscosities decrease with
increasing temperature as expected, spanning values from a low 193 cP to a high of
o
o
6298 cP over the temperature range 80 C to 25 C. (By comparison, the starting
o
material [P14,6,6,6][Cl] exhibits a viscosity of 1824 cP at 25 C [21].) The change in
viscosity corresponds to a physical change from being honey-like in character at
10

room temperature to being light oil-like at elevated temperatures. It is also evident
that the ferricyanide IL’s viscosity is effectively independent of the preparation
method over the entire temperature range studied which is a strong indication that
the materials are either relatively pure, or, at worse, have very similar impurity
profiles. A closer inspection of the data also demonstrates that they exhibit slight
positive curvatures suggesting that the classical hydrodynamic Arrhenius theory
does not apply which is a common observation for ionic liquids and other glass
forming liquids [20]. In such cases, the temperature-viscosity relationship is likely to
adhere to an alternative free volume model, typically the Vogel-Tamman-Fulcher
expression is assume to be appropriate [20] (VTF, Equation 3).
B
ln  A 
T - T
o
(3)
The VTF theory is based on the concept of temperature-dependent free
volume, where A and B are constants and T0 is the temperature at structural arrest
which is generally considered to correspond to the glassification temperature of the
liquid [22 – 24].
3.2 Thermal analysis
A review of the literature indicates that the glass transition temperature (Tg) for
o
o
[P14,6,6,6][Cl] is -56 C whilst a value of -70 C has been reported for [P14,
6,6,6]4[Fe(CN)6]4- [18]. Table 1 shows the results of the DSC determinations of Tg for
each of the four ferricyanide IL samples reported here, and for a ferrocyanide sample
(synthesis by the 1 + 1 ACN + H2O method). The values for ferricyanide are quite
o
consistent between synthetic procedures (-54 to -51 C) again indicating the
11

samples’ purity (or similar impurity profiles). It is interesting to note that the Tg
o
observed here for [P14, 6,6,6]4[Fe(CN)6]4- is 14 C higher than that previously reported
[18] and is very similar to the ferricyanide values and to the simple chloride salt.
Table 1 here
3.3 Electrochemistry of [P14,6,6,6]3[Fe(CN)6]3- and [P14,6,6,6]4[Fe(CN)6]4- in molecular
solvent and in ionic liquid [P14,6,6,6][NTf2]
Fe(CN)6]3-/4- is one of the most studies redox couples and it is well know that
under appropriate conditions the one-electron transfer is electrochemically reversible
with formal potentials (Eo  E1/2) that depend on the composition of the supporting
electrolyte. For example, at neutral pH, Eo = 0.361 V vs. SHE (Standard Hydrogen
Electrode) while in 0.1 mol L-1 HCl Eo = 0.56 V vs. SHE. It is therefore expected that
the nature of the supporting electrolyte will strongly influence the electrochemical
behaviour of the electroactive ionic liquids reported here.
Figure 4 shows a cyclic voltammogram for [P14,6,6,6]3[Fe(CN)6]3- in ACN/ TBATFB
electrolyte solution along with the potential reference compound ferrocene (Fc). It is
known that the formal potential for Fc/Fc+ is 0.64 V vs. SHE and is effectively
independent of the supporting medium.
Figure 4 here
In this system, the reversible Fc/Fc+ redox E1/2 occurs at ca. 0.180 V vs. Pt,
consistent with what is expected for the Fc/Fc+ couple. The other redox process,
although unusually distorted, with E1/2 of ca. -1.15 V vs. Pt, can be assigned to the
Fe(CN)6]3-/4- redox couple. It is immediately evident that under these conditions the
12

redox potential for this couple has shifted to extremely negative potentials, relative to
known aqueous systems. This can be seen in more detail in Figure 5 which shows
cyclic voltammograms as a function of potential sweep rate (). These
voltammograms also clearly indicate that the peak-to-peak separation increases with
increasing potential sweep rate, from 90 mV to 140 mV, upon going from 5 mV/s to
200 mV/s.
Figure 5 here
Although this behaviour might be attributed to slow heterogeneous electron transfer
kinetics, it is most likely due to uncompensated solution resistance (iR) within the cell
since the potentiostat does not have an IR correction facility. Also evident in the
voltammograms are the anodic-to-catholic peak-current ratios (Ipa/Ipc) at each
potential sweep-rate that are a consistent 1.0, which, unambiguously, indicates
electrochemical reversibility of [P14,6,6,6]3[Fe(CN)6]3 and very similar diffusion
coefficients for both the oxidised and deuced forms. Not unexpectedly, identical
electrochemical behaviour was observed for the [P14,6,6,6]4[Fe(CN)6]4 material in
ACN/TBATFB electrolyte solution which can be seen in Figure 6 where the slight
shift in E1/2 is probably due to variability of the Pt quasi-reference electrode.
Figure 6 here
For both systems in ACN electrolyte, peak currents are seen to increase linearly
½
with increasing  , behaviour in keeping with diffusion-controlled electrochemistry as
½
expected. Figure 7 shows linear Randles-Sevcik plots of Ipa and Ipc vs.  , the slopes
13

of which provide the mass transfer diffusion coefficients, D, for both redox forms
according to the Randles-Sevcik equation (at 298 K), Equation 4.
Ip = 2.69x105 n3/2 A D1/2 C 1/2
(4)
where C is the concentration of the electroactive spaces and n is the number of
electrons involved in the redox process. A D-value of 4.0 x 10-6 cm2 s-1 was obtained
for the ferricyanide form which is consistent with the molecular solvent’s low
viscosity. In addition, the near equivalence of the slopes for Ipa and Ipc suggest near
equivalent mass transport rates for both redox forms.
Figure 7 here
The electrochemistry of [P14,6,6,6]3[Fe(CN)6]3- and [P14,6,6,6]4[Fe(CN)6]4- in
[P14,6,6,6][NTf2] ionic liquid is presented in Figures 8 and 9 where identical
thermodynamic (E1/2) behaviour between both samples, and between both “solvent”
systems, is immediately evident. This equivalence indicates that the nature of
“solvation” of the Fe(CN)6]3-/4 species is similar under these highly-different
conditions. In addition, the equivalence of anodic and cathodic peak currents (i.e.
Ipa/Ipc 1.0) is indicative of electrochemical reversibility and near equivalence of the
diffusion coefficients of the respective redox couples.
Figures 8 & 9
14

There are, however, two significant dissimilarities with the behaviour observed in
molecular solvent. Firstly, as expected, the peak-current magnitudes are at least an
order of magnitude lower in the IL due to the viscosity effect. This effect can be seen
in the Randles-Sevcik plots for both materials that are shown in Figures 10 and 11
from which the corresponding D-values can be obtained. For [P14,6,6,6]3[Fe(CN)6]3- a
value for D of 1.0 x 10-8 cm2 s-1 was obtained whereas for a slightly larger value of
2.0 x 10-8 cm2 s-1 was obtained for [P14,6,6,6]4[Fe(CN)6]4-. Secondly, the peak-to-peak
separations (Ep, 150 mV to 460 mV upon going from 5 to 200 mV/s) are much
larger in the ILs. In the absence of IR correction facility, the larger Ep can, for the
most part at least, be attributed to uncompensated solution resistant although the
deleterious effect of high viscosity on electron transfer dynamics cannot be totally
discounted [25].
Figures 10 & 11 here
In order to understands physiochemical origin of the very negative redox potential
for the [P14,6,6,6]3/4[Fe(CN)6]3-/4] species work by Gutmann et al. [26-28] and Pletcher
et al. [29] are highly informative. These workers studied the electrochemistry of
Fe(CN)6]3-/4 in aprotic organic solvent and found the formal potentials to be ca. 1.0 -
1.3 V more negative in the aprotic solvents compared to aqueous values. The
Authors speculated that the difference arose from a strong stabilisation of the Fe2+
complex by water due to the donor properties of the nitrogen in [Fe(CN)6]4- being
much stronger than in [Fe(CN)6]3- as the Fe2+ is a weaker electron acceptor and also
a much stronger π-electron donor. Pletcher et al. [29] further considered the
following redox reaction involving a hydration / dehydration step (Equation 5)
15

4-
Fe(CN)6(H2O)p3- + (q - p) H2O + e- Fe(CN)6(H2O)q
(5)
where q > p. If diffusion and activity coefficients are ignored, the linear behaviour can
be accounted for using Nernst-like free energy function (Equation 6);
ΔEo = (Eo)solvent+water - (Eo)solvent = RT/F{log(KII/KIII) + (q-p) log[H2O]}
(6)
where KII and KIII are the overall stability constants for the formation of the
ferrocyanide/water and ferricyanide/water complexes, respectively.
In order to examine the effect of H2O on the electrochemistry of hydrophobic
[P14,6,6,6]3/4[Fe(CN)6]3-/4 ILs, aqueous solutions were prepared using the CTAB
surfactant in KCl/PBS buffer. A cyclic voltammogram of [P14,6,6,6]4[Fe(CN)6]4- in this
medium is shown in Figure 12 where classical 1-electron reversible electrochemistry
centred at ca. -0.125 V s. Pt is evident which is at least 1 V more positive than in
either the ACN or the IL environments which is entirely consistent with Pletcher et
al.’s [29] observations.
Figure 12 here
3.4 Stability of [P14,6,6,6]3[Fe(CN)6]3- and [P14, 6,6,6]4[Fe(CN)6]4-
It is well known that ferricyanides are prone to the adventitious formation of
Prussian blue (PB, ferric hexacyanoferrate, FeIII [FeII(CN)6]3) which, in the
4
appropriate context, is useful; for example, as a surface-immobilised redox catalyst
for H2O2 reduction in oxidase based biosensors such as the Glucomen Day glucose
biosensor [30]). The material synthesised here are not immune to PB formation and
16

its deposition on surfaces, although not investigated here, an ionic liquid form of PB
may have useful electrochemical properties and is currently the subject of further
investigation.
TGA was used to investigate the thermal stability of the Fe(CN)6]3-/4- ionic liquids
o
(and [P6,6,6,14][Cl] for comparative purposes) up to 800 C. A number of samples
o
exhibited weight loss at temperatures below 75 C which is likely to be the loss of
residual solvent or low molecular weight neutral contaminants. Overall, both
Fe(CN)6]3-/4- liquids exhibited excellent thermal stabilities with the materials
o
remaining stable up to at least 400 C. Onset decomposition temperatures, Tonset
,
defined as the temperature at which decomposition begins to occur, are shown in
Table 1. It is clear that the phosphonium cation of [P14,6,6,6][Cl] decomposes at 425
oC which compares favourably with the Tonset for the two Fe(CN)6]3-/4- ionic liquids
o
o
samples, i.e. 413 C and 431 C for ferric- and ferro- ILs, respectively. For both
o
Fe(CN)6]3-/4- samples, a second decomposition process was observed at 458 C
which, being absent for [P14,6,6,6][Cl], can be unambiguously attributed to the Fe(CN)6
moiety. Sun et al. [31] reported similar behaviour in the TGA traces for phosphonium
ILs based on organophosphate and organophosphinate anions.
4. Conclusions
The simple examples of functional electro-materials, based on redox active ionic
liquids, shown here, demonstrate the possibility of creating new potentially useful
materials with unique properties from easily available starting materials. These
examples of building electrochemical functionality into ionic liquid structure are
relatively straightforward from a synthetic perspective, but the likelihood of success
(obtaining a RTIL) depends entirely on the nature of the anion and cation, therefore
17

restricting synthesis to commercially available ions will inevitably severely limit the
functional materials obtained. Notwithstanding, concerning the materials
demonstrated here, the simple reversible electrochemistry, and their stability, bodes
well for their potential future application is energy storage devices (e.g. batteries and
hybrid capacitors), sensors, redox catalysis and bio-electrochemistry.
5. Acknowledgments
L. G. would like to thank the Department of Education and Learning in Northern
Ireland for a PhD Studentship. APD and GGW thank The Royal Society of London
for an International Collaboration Grant. And all the Authors thank Dr Paula Douglas
at the Polymer Processing Research Centre within QUB for performing the thermal
analysis.
18

6. References
1. J. S., Wilkes, M. J. Zaworotko, Chem. Commun. 13 (1992) 965 - 967.
2. F. H. Hurley, T. P. Wier, Jr., US Patent 2,446,349 (1948).
3. H. L.Chum, V. R. Koch, L. L. Miller, R. A. Osteryoung, J. Am. Chem. Soc. 91 (1975) 3264
- 3265.
4. T. Welton, Chem. Rev. 99 (1999) 2071 - 2083.
5. J. Holbrey, K. R. Seddon, Clean Products and Processes 1 (1999) 223-236.
6. P. Wasserscheid, T.Welton; Ionic Liquids in Synthesis; Wiley, VCH: Weinheim, 2002.
7. M. J. Earle, S. P. Katdare, K. R. Seddon, Org. Lett., 6 (2004 707-710
8. J. H. Davis, Chem Lett. 33 (2004) 1072 -1077.
9. E. D. Bates, R. D. Mayton, I. Ntai, J. H. Davis, J. Am. Chem. Soc. 124 (2002) 926-927.
10. M. E. Williams, R.W. Murray, J. Phys Chem. B. 103 (1999) 10221-10227.
11. H, Masui, R. W. Murray, Inorg. Chem. 36 (1997) 5118 - 5126.
12. P. K. Bhowmik, H, Haesook, J. J. Cebe, R. A. Burchett, Liquid Crystals, 30 (2003) 1433 -
1440.
13. Y. Gao, B. Twamley, J. M. Shreeve, Inorg. Chem. 43 (2004) 3406 - 3412.
14. M. G. Kang, K. S. Ryu, S. H. Chanh, N. G. Park, ETRI 26 (2004) 26 647- 652.
15. X. E. Wu, L. Ma, M.X. Ding, L.X. Gao, Synlett 4 (2005) 607-610.
16, V. Strehmel, H. Rexhausen, P. Strauch, Tet. Lett. 49 (2008) 7143-7145.
17. P. Bonhote, A. P.Dias, N. Papageorgiou, K. Kalyanasundaram, M. Gratzel, Inorg. Chem.
35 (1996) 1168–1178.
18. R. E. D. Sesto, C. Corley, A. Robertson, J. S. Wilkes, J. Organomet. Chem. 690 (2005),
2536-2542.
19 K. R. Seddon, A., Stark, M. J.Torres Pure Appl. Chem. 72 (2000) 2275-2287.
20 K. Paduszyński, U. Domańska, J. Chem. Inf. Model 54 (2014) 1311-1324
21. K. J. Fraser, D. R. MacFarlane, Aust. J. Chem., 62 (2009) 309–321.
22. H. Vogel Phys. Z., 22 (1921) 22, 645
19

23. G. Tamman, G. Hesse, Z. Anorg. Allg. Chem. 156 (1926) 245.
24. L. Wang J. Phys. Condens. Matter, 24 (2012) 155103.
25. C. L. Lagrost, D. Carrié, M. Vaultier, P. Hapiot, J. Phys. Chem. A, 107 (2003) 745 – 752.
26. V. Gutmann, G. Gritzner, K .Danksagmüller, Inorg. Chim. Acta. 17 (1976), 81- 86.
27. G. Gritzner, K. Danksagmüller, V.Gutmann; J. Electroanal. Chem. 72 (1976) 177-185.
28. G.Gritzner, D K.anksagmüller, V.Gutmann, J. Electroanal. Chem. 90 (1978) 203- 210.
29. D. Pletcher, R. E.Noftle, J. Electroanal. Chem. 293 (1990) 273-277.
30. F. Lucarelli, F Ricci , F, Caprio, F. Valgimigli, C. Scuffi, D. Moscone, G. Palleschi, J.
Diabetes Sci. Technol. 6 (2012) 1172-81.
31. J. Sun, P. C. Howlett, D. R. MacFarlane, J. Lin, M. Forsyth, Electrochim. Acta, 54
(2008) 254-260.
20

Table/Figures Legends
Table 1 Glass transition and thermal degradation temperatures of the IL samples as
recorded by DSC and TGA, respectively.
Sample
Synthesis
Tg (oC)
Tonset (oC)
T2nd
[P14,6,6,6]3[Fe(CN)6]3-
[P14,6,6,6]3[Fe(CN)6]3-
[P14,6,6,6]3[Fe(CN)6]3-
[P14,6,6,6]3[Fe(CN)6]3-
ACN / water (1:1)
ACN / water (1:2)
DMF / water (1:1)
Acetone / water (1:1)
-51
413
458
-54
-
-
-
-
-
-
-53
-54
[P14,6,6,6]4[Fe(CN)6]
ACN / water (1+1)
-56
-70*
-
431
425
458
-
[P14,6,6,6][NTf2]
[P14,6,6,6][Cl]
DCM/ water
-
-56*
* [18]
21

Figure 1 Structures of first stable room temperature ionic liquids [1].
Figure 2 Structure of [P14,6,6,6]3[Fe(CN)6]3/4-
.
Figure 3 Collation of Arrhenius plots for [P14,6,6,6]3[Fe(CN)6]3- and [P6,6,6,14]4[Fe(CN)6]4-
samples.
Figure 4 A cyclic voltammogram, recorded at a potential sweep rate of 100 mV/s, of
1.0 x 10-2 mol/L [P14,6,6,6]3[Fe(CN)6]3- and ferrocene in acetonitrile containing 0.1
mol/L TBATFB.
Figure 5 Cyclic voltammograms, recorded at the potential sweep rates indicated, of
1.0 x 10-2 mol/L [P14,6,6,6]3[Fe(CN)6]3- in acetonitrile containing 0.1 mol/L TBATFB.
Figure 6 Cyclic voltammograms, recorded at the potential sweep rates indicated, of
1.0 x 10-2 mol/L [P14,6,6,6]4[Fe(CN)6]4- in acetonitrile containing 0.1 mol/L TBATFB.
Figure 7 Randles-Sevcik plots for the reduction and re-oxidation of 1.0 x 10-2 mol/L
[P14,6,6,6]3 [Fe(CN)6]3- in acetonitrile containing 0.1 mol/L TBATFB.
Figure 8 Cyclic voltammograms, recorded at the potential sweep rates indicated, of
1.0 x 10-2 mol/L [P14,6,6,6]3[Fe(CN)6]3-in [P6,6,6,14][NTf2].
Figure 9 Cyclic voltammograms, recorded at the potential sweep rates indicated, of
1.0 x 10-2 mol/L [P14,6,6,6]4[Fe(CN)6]4-in [P6,6,6,14][NTf2].
Figure 10 Randles-Sevcik plots for the reduction and re-oxidation of [P14,6,6,6 3
]
[Fe(CN)6]3- in [P6,6,6,14][NTf2].
Figure 11 Randles-Sevcik plots for the reduction and re-oxidation of [P14,6,6,6 4
]
[Fe(CN)6]4- in [P6,6,6,14][NTf2].
22

Figure 12 A cyclic voltammogram of 4.6 x 10-3 mol/L [P14,6,6,6]4 [Fe(CN)6]4- recorded at
a potential sweep rate of 10 mV/s in 0.1 M KCl in 5 % (w/v) CTAB and 0.1 mol/L
PBS.
23