Voltammetric investigation of new boronic ester-substituted triphenylamines-based redox receptors in solution and attached to an electrode surface. Effects of F- as an anionic guest

Article abstract of DOI:10.1016/S0013-4686(01)00535-7

Triphenylamines mono-, di- and trisubstituted by boronic ester unit(s) were designed as powerful redox-active receptors for Lewis hard bases like fluoride anion. Their voltammetric behaviour was found to be dramatically changed upon the addition of this halide. Depending on the degree of substitution of triphenylamines, the binding of F^- to the boron atom led to the appearance of one to three new redox system(s). The binding constants determined for their neutral form ranged from 1.0×10² to 4.0×10² and were dramatically enhanced upon their oxidation into radical cation (3.0×10⁵-1.6×10⁷). The fixation of such electroactive compounds to the electrode surface has been achieved from the anodic oxidation of a vinyl-substituted bipodal receptor. The polymer films showed a reversible and stable response in a dried organic medium. Unfortunately, the voltammetric changes indicative of a complexation phenomenon were not observed in the presence of F^- and only a degradation of the film electroactivity was noticed.

Full text of DOI:10.1016/S0013-4686(01)00535-7

Electrochimica Acta 46 (2001) 3421–3429
www.elsevier.com/locate/electacta
Voltammetric investigation of new boronic ester-substituted
triphenylamines-based redox receptors in solution and
attached to an electrode surface. Effects of F− as an
anionic guest
M. Nicolas, B. Fabre *, J. Simonet
Laboratoire de Synthe`se et Electrosynthe`se Organiques (Unite´ Mixte de Recherche no 6510 associe´e au CNRS),
Uni6ersite´ de Rennes 1, Campus de Beaulieu, 35042 Rennes Cedex, France
Received 13 February 2001; received in revised form 26 April 2001
Abstract
Triphenylamines mono-, di- and trisubstituted by boronic ester unit(s) were designed as powerful redox-active
receptors for Lewis hard bases like fluoride anion. Their voltammetric behaviour was found to be dramatically
changed upon the addition of this halide. Depending on the degree of substitution of triphenylamines, the binding of
F− to the boron atom led to the appearance of one to three new redox system(s). The binding constants determined
for their neutral form ranged from 1.0×102 to 4.0×102 and were dramatically enhanced upon their oxidation into
radical cation (3.0×105–1.6×107). The fixation of such electroactive compounds to the electrode surface has been
achieved from the anodic oxidation of a vinyl-substituted bipodal receptor. The polymer films showed a reversible
and stable response in a dried organic medium. Unfortunately, the voltammetric changes indicative of a complexation
phenomenon were not observed in the presence of F− and only a degradation of the film electroactivity was noticed.
© 2001 Elsevier Science Ltd. All rights reserved.
Keywords: Triphenylamine; Boronic ester; Fluoride; Sensor; Cyclic voltammetry
such as ferrocene [3–5], bipyridine metal complex [6,7]
and triphenylamines [8]. The anodic behaviour of these
moieties was found to be changed upon the addition of
F− to the electrolyte solution. In the case of tripheny-
lamines mono- (TPAB1), di- (TPAB2) and trisubstituted
(TPAB3) by a pinacolborane unit (Fig. 1) [8], the
binding of F− to the boron atom of the receptor led to
the appearance of one to three new voltammetric
peak(s) at less positive potentials assigned to the oxida-
tion of each complexed form into its radical cation.
Furthermore, the interaction was specific to F− because
Br− or Cl− had no effect on the response of the three
triphenylamines. The efficiency of such anion receptors
having been demonstrated, this full paper aimed at
providing some essential data on the binding event,
1. Introduction
Among the neutral anion receptors, electron-deficient
groups like boronic acids and esters can strongly bind
hard-base anions such as fluoride, resulting in specific
orbital changes from sp2- to the more stable sp3-hy-
bridised boron [1–4].
With the aim of developing electroactive systems
capable of recognizing electrochemically and specifi-
cally anionic substrates, some boronic acids and esters
have been covalently attached to redox-active units
* Corresponding author. Tel: +33-29-9281628; fax: +33-
29-9286732.
E-mail address: fabre@univ-rennes1.fr (B. Fabre).
0013-4686/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0013-4686(01)00535-7

3422
M. Nicolas et al. / Electrochimica Acta 46 (2001) 3421–3429
plexation/decomplexation process. Moreover, the at-
tachment of such functional units to an electrode
surface to take advantage of their complexing proper-
ties observed in solution is undertaken from the anodic
oxidation of the vinyl-substituted bipodal receptor
VTPAB2. Finally, the effect of a Lewis hard base like
fluoride on the voltammetric response of the resulting
functionalised polymer film is also presented.
2. Experimental
2.1. Synthesis of boronic ester-substituted
triphenylamines
Solvents were freshly distilled over Na or P2O5 prior
to use. 1H and 13C NMR spectra were recorded in
CDCl3 solution with a Bruker AC 300 spectrometer
using TMS as an internal reference. Melting points
were determined with an Electrothermal® melting point
apparatus. Silica gel (70–230 mesh) for column chro-
matography was purchased from Silicycle. Mass spectra
were measured with a Varian MAT 311 spectrometer.
The synthesis of different pinacol borane-substituted
triphenylamines TPAB1, TPAB2 and TPAB3 (Fig. 1)
has been described previously [8].
Fig. 1. Pinacol borane-substituted triphenylamines.
such as the equilibrium constants corresponding to the
complexation of F− by the neutral and oxidised forms
of the receptor, but also the reversibility of the com-
VTPAB2 was obtained in five steps from tripheny-
lamine (Scheme 1). In the first step, triphenylamine
(Acros, 99%) was converted into 4-acetoxytripheny-
lamine 1 according to the procedure of Staskun [9].
The reduction of the ketone into the corresponding
alcohol 2 was achieved with LiAlH4. In a 250 ml
three-necked flask under an argon flow, 4-acetoxyt-
riphenylamine (7.18 g, 25 mmol) dissolved in anhy-
drous THF (30 ml) was added dropwise to
a
well-stirred suspension of LiAlH4 (0.95 g, 25 mmol) in
THF (50 ml) maintained at 0 °C. Following the addi-
tion, the mixture was allowed to slowly warm to room
temperature and the progress of the reaction was fol-
lowed by thin-layer chromatography (eluent: hexanes/
ethyl acetate 1:1 (v/v)). Then, the solution was cooled
to 0°C and carefully neutralised with water and few
drops of aq. NaOH 15%. After extraction with diethyl
ether, the organic phase was separated, washed with
water, dried with MgSO4 and concentrated to dryness
under reduced pressure. The crude product was purified
by flash chromatography (eluent: hexanes/ethyl acetate
4:1) to give 1-(4-diphenylamino-phenyl)ethanol 2 (5.73
g, 19.8 mmol) as a white solid.
Yield: 79%; m.p. 102–103°C. 1H NMR 300 MHz
(CDCl3, l ppm): 1.52 (d, CH3, [3JHH=6.4], 3H); 2.13
(s, OH, 1H); 4.86 (q, CH, [3JHH=6.4], 1H); 7.00–7.30
(m, HPh, 14H). 13C NMR 75 MHz (l ppm): 24.98
(CH3); 70.02 (CHOHCH3); 124.01, 124.16 and 129.23
(CPhH); 122.76 and 126.44 (CPhH); 139.97
Scheme 1. Synthesis of the vinyl-substituted bipodal redox-ac-
tive receptor VTPAB2. Reagents and conditions: (i) acetic
anhydride, toluene, then addition of SnCl4, reflux, 1 h. [9]; (ii)
LiAlH4, anhydrous THF, O°C; (iii) Br2, CHCl3, −20°C; (iv)
n-BuLi, trimethyl borate, anhydrous diethyl ether, −78°C,
pinacol, 25°C; (6) POCl3, pyridine, reflux, \12 h.
(CPhCHOHCH3); 147.72 and 147.83 (NCPh). MS m/e:
’
289.1464 (M+ ). Calc. 289.1467.

M. Nicolas et al. / Electrochimica Acta 46 (2001) 3421–3429
3423
A
solution of 1-(4-diphenylaminophenyl)ethanol
139.70 (CPhCHOHCH3); 146.07 (NCPh); 150.16 (NCPh).
’
MS m/e: 523.3104 ([M−H2O]+ ). Calc. 523.3065.
(5.00 g, 17.3 mmol) in CHCl3 (60 ml) was cooled to
−20 °C with a CH3CN+CO2 bath under stirring. At
this temperature, the slow addition of two molar equiv.
of Br2 (1.77 ml, 34.6 mmol) turned the solution green.
The progress of the reaction was followed by thin-layer
chromatography (eluent: petroleum ether/ethyl acetate
1:1 (v/v)). After 2 h, the mixture was allowed to warm
gently to room temperature and extracted with water.
Then, the organic phase was separated, dried with
MgSO4 and concentrated to dryness under reduced
pressure. The crude product was purified by chro-
matography on silica gel (eluent: petroleum ether/ethyl
acetate 1:1 (v/v)) to give 1-(4-[bis-(4-bromophenyl)-
amino]-phenyl)ethanol 3 (6.60 g, 14.8 mmol) as a
brown oil.
In a 25 ml three-necked flask under an argon flow,
1-4-(bis-[4-(4,4,5,5-tetramethyl-[1,3,2]dioxaborolane-2-
yl)-phenyl]-amino)-phenyl ethanol (1.00 g, 1.85 mmol)
and POCl3 (0.17 ml, 1.85 mmol) were dissolved in
distilled pyridine (10 ml). The mixture was heated un-
der reflux for more than 12 h and the solution initially
yellow turned progressively brown. The progress of the
reaction was followed by thin-layer chromatography
using diethyl ether/petroleum ether (1:1 (v/v)) as the
eluent. After cooling to room temperature, the solution
was extracted with a water/diethyl ether mixture. The
aqueous phase was treated with ethyl acetate and then,
with dichloromethane. The combined organic phases
were dried with MgSO4 and the solvents were evapo-
rated under reduced pressure. The resulting crude
product was chromatographed on silica gel (eluent:
diethyl ether/petroleum ether 1:1 (v/v)) to give bis-[4-
(4,4,5,5-tetramethyl-[1,3,2]dioxaborolane-2-yl)-phenyl]-
(4-vinylphenyl) amine VTPAB2 (0.32 g, 0.61 mmol) as a
white solid.
Yield: 85%. 1H NMR 300 MHz (CDCl3, l ppm):
1.35 (d, CH3, [3JHH=6.5], 3H); 1.54 (s, OH, 1H); 4.29
(q, CH, [3JHH=6.5], 1H); 6.85 and 7.25 (AA%BB% sys-
tem, HPh, [3JHH=8.8], 8H); 6.93 and 7.12 (AA%BB%
system, HPh, [3JHH=8.5], 4H). 13C NMR 75 MHz (l
ppm): 23.94 (CH3); 76.63 (CHOHCH3); 115.42
(CPhBr); 124.42 and 127.30 (CPhH); 125.54 and 132.36
(CPhH); 139.65 (CPhCHOHCH3); 145.94 (NCPh); 146.56
(NCPh).
The conversion of the dibrominated triphenylamine
into its ‘pinacol borane’ (4,4,5,5-tetramethyl-1,3,2-diox-
aborolane)-substituted analogous was achieved as fol-
lows: 1-(4-[bis-(4-bromophenyl)-amino]-phenyl)ethanol
(6.50 g, 14.5 mmol) dissolved in anhydrous diethyl
Yield: 33%; m.p. 78–79 °C. 1H NMR 300 MHz
(CDCl3, l ppm): 1.26 (s, CH3, 24H); 5.11 (dd, Ha,
[3JHaHc=10.9, 3JHaHb=0.8], 1H); 5.59 (dd, Hb,
[2JHbHc=17.5, 3JHbHa=0.8], 1H); 6.59 (dd, Hc,
[2JHbHc=17.5, 3JHaHc=10.9], 1H); 6.97 and 7.35
(AA%BB% system, HPh, [3JHH=8.6], 4H); 6.99 and 7.60
(AA%BB% system, HPh, [3JHH=8.4], 8H). 13C NMR 75
MHz (l ppm): 25.28 (CCH3); 84.05 (CCH3); 113.22
(CHCH2); 123.34 and 136.33 (CPhH); 125.56 and
127.59 (CPhH); 133.49 (CPhCHCH2); 136.53 (CHCH2);
ether (50 ml) was cooled to −78°C with
a
CH3COCH3+CO2 bath under stirring. At this temper-
ature, four molar equiv. of n-BuLi (58 mmol) were
added dropwise. After 2 h, the solution was allowed to
warm gently to room temperature. After cooling to
−78 °C, trimethyl borate (6.47 ml, 58 mmol) was
added and the reaction mixture was stirred for 2 h.
Then, 4 equiv. of pinacol (6.85 g, 58 mmol) in anhy-
drous diethyl ether were added to the solution warmed
to room temperature. The mixture was stirred for more
than 12 h. The solution was then washed with water,
the organic phase was separated, dried with MgSO4,
and the solvent was evaporated under reduced pressure.
The crude product was purified by column chromatog-
raphy (eluent: diethyl ether/petroleum ether 1:1 (v/v))
to afford 1-4-(bis-[4-(4,4,5,5-tetramethyl-[1,3,2]dioxa-
borolane-2-yl)-phenyl]-amino)-phenyl ethanol 4 (1.71 g,
3.16 mmol) as a yellowish solid.
147.05 (NCPh); 150.33 (NCPh). MS m/e: 523.3052
’
(M+ ). Calc. 523.3065.
2.2. Electrochemical instrumentation and procedures
Tetra-n-butylammonium perchlorate Bu4NClO4 from
Fluka (puriss, electrochemical grade), triethylamine tri-
hydrofluoride Et3N·3 HF and chlorotrimethylsilane
Me3SiCl from Aldrich (98%) were used as received.
Acetonitrile (max. 50 ppm water) from Merck and
methylene chloride (anhydrous analytical grade) from
SDS were used without further purification and stored
under dry argon. All electrolytic solutions were dried in
situ over neutral alumina from Merck, previously acti-
vated at 450°C under vacuum for several hours. They
were thoroughly deaerated and kept under a positive
pressure of dry argon during each run.
Linear potential sweep cyclic and differential pulse
voltammetry experiments were performed with an Au-
tolab PGSTAT 20 potentiostat from Eco Chemie B.V.,
equipped with General Purpose Electrochemical System
GPES software (version 4.5 for Windows). The work-
ing electrode was a platinum disk (area: 0.8 mm2) and
Yield: 21.8%; m.p. 62–63 °C. 1H NMR 300 MHz
(CDCl3, l ppm): 1.13 (s, CH3, 24H); 1.24 (d, CH3,
[3JHH=6.4], 3H); 1.41 (s, OH, 1H); 4.17 (q, CH,
[3JHH=6.4], 1H); 6.78 and 6.99 (AA%BB% system, HPh
[3JHH=8.5], 4H); 6.85 and 7.47 (AA%BB% system, HPh
,
,
[3JHH=8.4], 8H). 13C NMR 75 MHz (l ppm): 23.91
(CH3); 24.89 (CCH3); 77.30 (CH); 83.67 (CCH3);
122.76 and 135.90 (CPhH); 125.35 and 127.15 (CPhH);

3424
M. Nicolas et al. / Electrochimica Acta 46 (2001) 3421–3429
3. Results and discussion
3.1. Voltammetric beha6iour of the boronic
ester-substituted triphenylamines
As expected, the electrochemical behaviour of
triphenylamines mono- and disubstituted by boronic
ester groups in CH3CN+0.1 M Bu4NClO4 is similar to
that obtained for other triphenylamines [10–13]. At 0.1
V s−1, TPAB1 and TPAB2 are irreversibly oxidised at
E
pa=0.69 and 0.74 V, respectively, and the backward
scan shows small cathodic peaks located at 0.44 and
0.54 V for TPAB1 and 0.47 and 0.56 V for TPAB2 (Fig.
2). A one-electron reversible process was observed at
E°%=0.655 V for TPAB1 and 0.705 V for TPAB2 when
the scan rate was increased to 5 V s−1. Moreover, the
ratio Ipa/61/2 (where Ipa is the anodic peak current
intensity and 6 is the potential scan rate) was found to
decrease with increasing 6 and Epa was positively
shifted of about 20 mV for a ten-fold increase in 6 (Fig.
3). For scan rates larger than 0.4 V s−1, Epa was
approximately constant. Thus, these results are consis-
tent with a second-order ECE mechanism in which the
unstable electrogenerated radical cation leads to a
dimer that is more easily oxidisable than the starting
material [13]. Experimentally, the dimerisation constant
kd of each triphenylamine can be determined from the
variation of Epa with 6 [14] (Eq. (1)).
F
6i
kd=0.80
(1)
RT C0
where 6i is the scan rate measured at the intersection
between the two linear parts of the Epa−log 6 experi-
mental diagram (Fig. 3) and C0 is the bulk concentra-
tion of the substituted triphenylamine. So, values of
2.6×103 l mol−1 s−1 and 2.4×103 l mol−1 s−1 were
extracted for TPAB1 and TPAB2, respectively. Com-
pared to those reported in the literature for other
substituted triphenylamines [12,15], it could be con-
cluded that the presence of electron-withdrawing
boronate groups increased the coupling probability be-
tween two radical cations, resulting in a larger dimerisa-
tion constant. Indeed, triphenylamines mono- or
disubstituted by an electron-donating methoxy group
were characterised by a small kd (lower than 6×10−1
l mol−1 s−1) whereas the presence of nitro groups led
Fig. 2. Cyclic voltammograms in CH3CN+0.1 M Bu4NClO4
of: (a) TPAB1 (2 mM); (b) TPAB2 (2 mM); and (c) TPAB3
(1.5 mM). Potential scan rate: 0.1 V s−1
.
to kd larger than 104 l mol−1 s−1
.
In contrast with the voltammetric behaviour of
TPAB1 and TPAB2, the tri-para-substituted tripheny-
lamine TPAB3 exhibited, whatever the scan rate, a
reversible oxidation step at E°%=0.72 V corresponding
to the formation of a stable radical cation (Fig. 2c).
Furthermore, the ratio Ipa/61/2 was roughly constant
within the scan rate range investigated (0.01–10 V s−1).
the counter electrode was a glassy carbon rod. All
potentials were relative to the system 0.1
AgNO3 ꢀ Ag in acetonitrile. Solution resistance was
compensated by positive feedback and the accuracy of
the corrections was tested with ferrocene at different
concentrations.
M

M. Nicolas et al. / Electrochimica Acta 46 (2001) 3421–3429
3425
consistent with a specific recognition of F− by TPABn
(with n ranging from 1 to 3).
3.2. Effect of F− on the 6oltammetric response of the
boronic ester-substituted triphenylamines
We have checked by cyclic voltammetry that the
different complexed species were reversibly oxidised, i.e.
the peak-to-peak separation was about 60–70 mV and
Ipa/Ipc:1. So, in these conditions, the fluoride com-
plexation by TPABn can be described by the various
equilibria depicted in Scheme 2. The F−-binding con-
stant Km corresponding to the neutral form of TPABn
was determined for each susbtituted triphenylamine
from DPV data (see Appendix A) and the calculated
values are summarised in Table 1. Moreover, the bind-
ing constant corresponding to the oxidised form of
TPABn, K% , was easily obtained from the ratio K% /Km
Owing to the electronic communication between the
boronic ester unit(s) and the redox center, it is expected
that the binding of a hard base like F− leads to the
modification of the electrochemical response of the
receptor. The effect of this halide has been analysed in
CH2Cl2 because of the ability of this solvent to stabilise
radical cations [16]. As clearly evidenced by differential
pulse voltammetry (DPV), the initial redox system as-
signed to the oxidation of the substituted tripheny-
lamine into its corresponding radical cation was
decreased in intensity upon the addition of F− while
one or several new peak(s) emerged at less positive
potentials (Fig. 4). Those were assigned to the oxida-
tion of each complexed form of triphenylamines and
their lower potential was explained by the stabilisation
of the radical cation upon the binding of F− to the
boron atom.
The presence of the pinacolborane group was essen-
tial for electrochemically sensing F− because the
voltammetric response of an unsubstituted tripheny-
lamine was found to be unmodified after the addition
of this halide. Moreover, as demonstrated in our pre-
liminary paper [8], no changes were noticed when F−
was replaced by Br− or Cl−. All these results are thus
m
m
expressed as a function of the formal potentials of
complexed and free forms (Eq. (2)).
K%
m=exp
Km
F
ꢁ
n
(E°m% −1−E°% )
(2)
m
RT
Expectedly, it appears that the complexes produced
between the TPABn radical cation and F− were about
103–5×104-fold more stable than those formed with
the neutral species. It must be pointed out that a
weaker enhancement binding upon electron transfer has
been reported for ferrocene boronic acid [4], which
indicates that TPABn is a more powerful F− redox-ac-
tive receptor.
Fig. 3. Ipa/61/2−6 and Epa−log 6 plots for (a, c) TPAB1; and (b, d) TPAB2 at 2 mM in CH3CN+0.1 M Bu4NClO4.

3426
M. Nicolas et al. / Electrochimica Acta 46 (2001) 3421–3429
3.3. Re6ersibility of the fluoride complexation/
decomplexation process
Generally, the development of a sensory device for
long-term uses requires that the properties of the sens-
ing element are not irremediably changed after the
binding event, i.e. the original signal is recovered with-
out significant loss of intensity. In the contrary case, the
system can function only if a preferentially mild and
fast method yields the original signal. Taking into
account the latter feature, chlorotrimethylsilane
Me3SiCl seemed to be an efficient reagent to cleave the
boron-fluoride bond of complexed TPABn. As shown in
Fig. 5 for TPAB1, the redox system attributed to the
complexed form totally disappeared after the addition
of Me3SiCl and the initial peak was restored with the
same intensity but shifted 50 mV towards less positive
potentials. The reason for such a shifting is not eluci-
dated but it is possible that the silane derivative also
interacts with nitrogen of the substituted tripheny-
lamine [17] resulting in the modification of its anodic
behaviour. Similarly to TPAB1, the DPV data obtained
with TPAB2 and TPAB3 were consistent with the con-
version of the various complexed forms into their free
form after the addition of Me3SiCl. Thus, all these
results demonstrate that the decomplexation process
can be chemically achieved without significantly lower-
ing the redox properties of the receptor.
3.4. Anodic electropolymerisation of VTPAB2. Effect
of fluoride
In the light of the promising sensing potentialities
exhibited by TPABn in homogeneous phase, the immo-
bilisation of such a redox-active receptor onto an elec-
trode surface has been considered, using the
vinyl-substituted bipodal triphenylamine VTPAB2 as
the starting precursor. The cyclic voltammogram of the
latter in acetonitrile medium showed two irreversible
anodic peaks at 0.60 and 0.74 V (0.1 V s−1) (Fig. 6a).
Repetitive potential cycling between 0 and 1.0 V led to
the formation of an electroactive polymer film
poly(VTPAB2), as revealed by the increase of the sec-
ond anodic peak and its corresponding cathodic com-
ponent emerging on the reverse scan. After several
cycles, the modified electrode was transferred to a
monomer-free electrolyte solution and showed a per-
fectly reversible redox system at 0.61 V (Fig. 6b). As
already reported for other vinyl-substituted electroac-
tive compounds [18–20], the polymerisation mechanism
Fig. 4. Three-dimensional view of differential pulse voltam-
mograms DPV in CH2Cl2+0.1 M Bu4NClO4 of: (a) TPAB1;
(b) TPAB2; and (c) TPAB3 at 1 mM, as a function of the
Et3N·3 HF concentration (from 0 to 5 mM). DPV parameters:
modulation time, 0.1 s; interval time, 1 s; modulation ampli-
tude, 50 mV.
has chain propagation character and yields
a
polystyrene-like polymer. So, the material electroactiv-
ity is due to the electron hopping between the neutral
triphenylamine and its radical cation species. The films
could be repeatedly cycled without apparent decrease of
their electroactivity (Fig. 6b), provided that the elec-
Scheme 2. General square scheme for the F− complexation by
boronic ester-substituted triphenylamines. The formal poten-
tials E°m% −1 and E°% were extracted from DPV of Fig. 4.
m

M. Nicolas et al. / Electrochimica Acta 46 (2001) 3421–3429
3427
Table 1
Binding constants corresponding to the complexation of F− by the neutral and oxidised forms of three substituted triphenylamines
K1
K%
1
K2
K%
2
K3
K%
3
TPAB1
TPAB2
TPAB3
2.4×102
2.8×102
2.1×102
1.6×107
1.2×107
0.9×107
–
–
–
–
–
–
3.9×102
1.6×102
1.1×106
3.0×105
1.0×102
3.4×105
trolytic medium was thoroughly dried with activated
neutral alumina. Under these conditions, the peak-to-
peak separation (DEp) was close to 20 mV and the
half-height peak width (DEp/2) was 300 mV at 0.05
V s−1. Theoretically, values of 0 and 90 mV are ex-
pected for surface-immobilised redox species with a
monoelectronic Nernstian-type reaction [21]. The larger
DEp/2 value would be indicative of strong repulsive
interactions between electroactive centers [22]. More-
over, the fixation of the receptor to the electrode sur-
face was also proved by the proportionality between Ipa
and the scan rate up to 1 V s−1. When studied in
CH2Cl2, poly(VTPAB2) was characterised by a less
reversible and more distorded system (DEp:100 mV at
0.05 V s−1).
The effect of F− on the electrochemical behaviour of
poly(VTPAB2) is illustrated in Fig. 7. In dried CH3CN,
the anodic peak was greatly decreased in size and
broadened upon the addition of this halide. The film
electroactivity was even totally lost from a certain F−
concentration, for example 5 mM with a film elec-
trosynthesised using 150 mC cm−2, and not subse-
quently restored in a solution without fluoride. The
same trend was observed for other polymer thicknesses
and in previously dried CH2Cl2. Consequently, the
voltammetric changes caused by F− in homogeneous
phase (namely, decrease of the initial system together
with the emergence of new redox system(s)) were not
seen in heterogeneous phase. As the degradation of the
electroactive properties was irreversible, it could be
concluded that poly(VTPAB2) was not suitable for
electrochemically sensing F−. The strong F−-induced
electroactivity decrease was thought to be ascribed to a
film conductivity lowering caused by the large repulsive
interactions between the sterically hindered tripheny-
lamine units. Nevertheless, a partial solubilisation of
the F−-complexed film could also account for the
voltammetric data.
Fig. 5. DPV in CH2Cl2+0.1 M Bu4NClO4 of TPAB1 (1 mM)
before (i) and after addition of 5 mM Et3N·3 HF (ii) and then
10 mM Me3SiCl (iii).
4. Conclusions
Fig. 6. (a) Potentiodynamical electropolymerisation of
VTPAB2 at 5 mM in CH3CN+0.1 M Bu4NClO4 and (b)
response of the resulting polymer film in the same electrolytic
solution; first cycle (—) and tenth cycle (— —). Potential
As demonstrated in this paper, the sensing properties
of a redox-active receptor can be greatly diminished
after its immobilisation onto an electrode surface. In-
deed, the remarkable voltammetric effects observed
scan rate: 50 mV s−1
.

3428
M. Nicolas et al. / Electrochimica Acta 46 (2001) 3421–3429
can be determined from the DPV anodic peak currents
Ipa of each redox system, assuming that the diffusion
coefficients of the free and complexed forms are nearly
equal.
For TPAB1, the F− binding constant K1 is given by
[TPAB1−(F−)]
[TPAB1][F−]
K1=
(A1)
where [TPAB1−(F−)], [TPAB1] and [F−] correspond
to the F−-complexed TPAB1, free TPAB1 and free
fluoride concentrations, respectively (the standard con-
centration used is 1 mol dm−3).
Taking into account the mass conservation for
fluoride,
Fig. 7. DPV of a poly(VTPAB2) film (electropolymerisation
charge, 150 mC cm−2) in CH3CN+0.1 M Bu4NClO4 without
(i) and with 0.1 mM (ii), 0.5 mM (iii), 1.0 mM (iv) and 2.0 mM
(v) of Et3N·3 HF.
[F−]tot=[F−]+[TPAB1−(F−)]
(A2)
with TPABn in the presence of fluoride are changed
into a total and irreversible degradation of the receptor
electroactivity with poly(VTPAB2).
where [F−]tot is the total fluoride concentration, K1 can
be re-written
It seems obvious that the recognition pathway in
heterogeneous phase is dependent on a large number of
not easily controllable parameters, such as the analyte
diffusion and the transport of electrons through the
film as well as on the stability of charge carriers and the
polymer structure. In contrast, the situation is consider-
ably simplified when both the receptor and the anionic
guest are in solution. Generally, such a feature is not
raised in the literature and most of concerned papers
have dealt with recognition events occurring in homo-
geneous phase. In our case, different strategies could be
applied in order that the immobilised receptor retains
its original characteristics, including the chemical poly-
merisation of VTPAB2 with a,a%-azoisobutyronitrile
(AIBN) or benzoyl peroxide [23] (influence on the
cross-linking and the polymerisation degree) or intro-
duction of the boronic ester-substituted triphenylamine
units into a backbone of conjugated polymer [24].
[TPAB1−(F−)]
K1=
(A3)
[TPAB1]([F−]tot−[TPAB1−(F−)])
or
1
[TPAB1]
K1 [TPAB1−(F−)]
=
[F−]tot−[TPAB1]
(A4)
This parameter can be expressed as a function of the
DPV anodic peak currents Ipa assigned to the free and
complexed TPAB1 forms (Eq. (A5)).
1
Ipa(TPAB1)
K1 Ipa(TPAB1−(F−))
Ipa(TPAB1)
I0pa(TPAB1)
=
[F−]tot−[TPAB1]0
(A5)
where I0pa(TPAB1) is the anodic peak current recorded
in the absence of fluoride and [TPAB1]0 is the initial
TPAB1 concentration.
For TPAB2, the constants attributed to the binding
of one and then two fluoride anions can be expressed as
follows:
Acknowledgements
Financial support from the ‘Re´gion Bretagne’ (fel-
lowship to M.N.) is gratefully acknowledged. J.-M.
Chapuzet and J. Lessard (Laboratoire d’Electrochimie
Organique, Centre de Recherche en Electrochimie et
Electrocatalyse, Universite´ de Sherbrooke, Canada) are
thanked for their help at the beginning of this work.
The authors wish to thank the Universite´ de Rennes 1
and CNRS (Contract UMR 6510) for technical and
financial support. They are also grateful to Dr M.
Vaultier for fruitful discussions.
[TPAB2−(F−)]
K1=
K2=
(A6)
(A7)
[TPAB2][F−]
[TPAB2−(F−)2]
[TPAB2−(F−)][F−]
where [TPAB2], [TPAB2−(F−)] and [TPAB2−(F−)2]
are the free TPAB2, one and two F−-complexed TPAB2
concentrations, respectively. Considering the mass con-
servation for fluoride,
Appendix A
1
[F−]tot=[F−]+[TPAB2−(F−)]+ [TPAB2−(F−)2]
The binding constants Km corresponding to the free
and complexed neutral forms of TPABn (with n=1–3)
2
(A8)

M. Nicolas et al. / Electrochimica Acta 46 (2001) 3421–3429
3429
it can be easily established that
References
1
Ipa(TPAB2)
[1] T.D. James, K.R.A. Samankumara Sandanayake, S.
Shinkai, Angew. Chem. Int. Ed. Engl. 35 (1996) 1910.
[2] C.R. Cooper, N. Spencer, T.D. James, Chem. Commun.
(1998) 1365.
[3] H. Yamamoto, A. Ori, K. Ueda, C. Dusemund, S.
Shinkai, J. Chem. Soc. Chem. Commun. (1996) 407.
[4] C. Dusemund, K.R.A. Samankumara Sandanayake, S.
Shinkai, J. Chem. Soc. Chem. Commun. (1995) 333.
[5] A.N.J. Moore, D.D.M. Wayner, Can. J. Chem. 77 (1999)
681.
[6] T. Mizuno, M. Takeuchi, I. Hamachi, K. Nakashima, S.
Shinkai, J. Chem. Soc. Perkin Trans. 2 (1998) 2281.
[7] M. Nicolas, B. Fabre, J. Simonet, Electrochim. Acta 46
(2001) 1179.
[8] M. Nicolas, B. Fabre, J.-M. Chapuzet, J. Lessard, J.
Simonet, J. Electroanal. Chem. 482 (2000) 211.
[9] B. Staskun, J. Org. Chem. 33 (1968) 3031.
[10] R.F. Nelson, S.W. Feldberg, J. Phys. Chem. 73 (1969)
2623.
[11] L.S. Marcoux, R.N. Adams, S.W. Feldberg, J. Phys.
Chem. 73 (1969) 2611.
[12] S.C. Creason, J. Wheeler, R.F. Nelson, J. Org. Chem. 37
(1972) 4440.
[13] E.T. Seo, R.F. Nelson, J.M. Fritsch, L.S. Marcoux, D.W.
Leedy, R.N. Adams, J. Am. Chem. Soc. 88 (1966) 3498.
[14] C.P. Andrieux, L. Nadjo, J.-M. Save´ant, J. Electroanal.
Chem. 42 (1973) 223.
=
[F−]tot−[TPAB2]0
K1 Ipa(TPAB2−(F−))
IpaTPAB2
I0pa(TPAB2)
1 I (TPAB −(F−) )
ꢂ
ꢃ
pa
2
2
×
1+
(A9)
2 Ipa(TPAB2−(F−))
and
K2 Ipa(TPAB2−(F−)2)Ipa(TPAB2)
=
(A10)
K1
I2pa(TPAB2−(F−))
TPAB3 is characterised by three binding constants
(Eqs. (A11), (A12) and (A13))
[TPAB3−(F−)]
[TPAB3][F−]
[TPAB3−(F−)2]
[TPAB3−(F−)][F−]
[TPAB3−(F−)3]
[TPAB3(F−)2][F−]
K1=
K2=
K3=
(A11)
(A12)
(A13)
with [TPAB3], [TPAB3−(F−)], [TPAB3−(F−)2] and
[TPAB3−(F−)3] are the free TPAB3, one, two and
three F−-complexed TPAB3 concentrations, respec-
tively. The value of the first binding constant is ex-
tracted from Eq. (A14) and the two following ones are
then easily deduced.
[15] R.F. Nelson, R.N. Adams, J. Am. Chem. Soc. 90 (1968)
3925.
[16] C.K. Mann, in: A.J. Bard (Ed.), Electroanalytical Chem-
istry, vol. 3, Marcel Dekker, New York, 1969, p. 57.
[17] E.W. Colvin, Silicon in Organic Synthesis, Butterworths,
London, 1981.
1
[F−]tot=[F−]+[TPAB3−(F−)]+ [TPAB3−(F−)2]
2
1
+
[TPAB3−(F−)3]
(A14)
3
1
Ipa(TPAB3)
[18] J.-C. Moutet, J. Electroanal. Chem. 161 (1984) 181.
[19] F. Bedioui, J. Devynck, C. Bied-Charreton, Acc. Chem.
Res. 28 (1995) 30.
[20] M.R. Van De Mark, L.L. Miller, J. Am. Chem. Soc. 100
(1978) 3223.
[21] A.J. Bard, L.R. Faulkner, Electrochemical Methods.
Fundamentals and Applications, Wiley, New York, 1980,
p. 522.
[22] E. Laviron, in: A.J. Bard (Ed.), Electroanalytical Chem-
istry, vol. 12, Marcel Dekker, New York, 1982.
[23] R.G. Compton, M.E. Laing, A. Ledwith, I.I. Abu-Ab-
doun, J. Appl. Electrochem. 18 (1988) 431.
[24] W.-L. Yu, J. Pei, W. Huang, A.J. Heeger, Chem. Com-
mun. (2000) 681.
=
[F−]tot−[TPAB3]0
K1 Ipa(TPAB3−(F−))
I (TPAB )
1 Ipa(TPAB3−(F−)2)
2 Ipa(TPAB3−(F−))
ꢂ
pa
3
×
+
1+
I0pa(TPAB3)
1 I (TPAB −(F−) )
ꢃ
pa
3
3
(A15)
(A16)
(A17)
3 Ipa(TPAB3−(F−))
K2 Ipa(TPAB3−(F−)2)Ipa(TPAB3)
=
K1
I2pa(TPAB3−(F−))
K3 Ipa(TPAB3−(F−)3)Ipa(TPAB3−(F−))
=
K2
I2pa(TPAB3−(F−)2)
.