Electrochemical investigation on the polycondensation kinetics of silicon alkoxides by functionalization of the silica network by redox species

Article abstract of DOI:10.1021/jp021872l

The sol-gel polycondensation of tetramethoxysilane has been followed for the first time by functionalization of the oligomeric silane species with a redox active ferrocene. Recording the decrease of the average diffusion coefficient of the mobile species brings information on the sol or gel state, as well as an easy insight of the polycondensation kinetics.

Full text of DOI:10.1021/jp021872l

J. Phys. Chem. B 2003, 107, 1321-1325
1321
Electrochemical Investigation on the Polycondensation Kinetics of Silicon Alkoxides by
Functionalization of the Silica Network by Redox Species
P. Audebert,* S. Sallard, and S. Sadki
Laboratoire de Photophysique et Photochimie Supramol e´ culaires et Macromol e´ culaires, Ecole Normale
t
Sup e´ rieure de Cachan, 61 AV. du P Wilson, 94235 CACHAN, Cedex France
ReceiVed: August 12, 2002; In Final Form: October 7, 2002
The sol-gel polycondensation of tetramethoxysilane has been followed for the first time by functionalization
of the oligomeric silane species with a redox active ferrocene. Recording the decrease of the average diffusion
coefficient of the mobile species brings information on the sol or gel state, as well as an easy insight of the
polycondensation kinetics.
Introduction
In a first step, we describe the synthesis of the adequately
functionalized ferrocene-siloxane. Then we show that the
In the course of our current investigations on the electro-
chemical response of sol-gel systems, we have previously
shown that in the case of zirconium propoxide sol-gel
polycondensation that electrochemistry could be a particularly
useful tool to investigate the size and the mobility of the species
formed, for example by introducing a small amount of an
electroactive compound bearing also the same functional group
electrochemical data throughout the course of TMOS polycon-
densation allows us to differentiate the hydrolysis, condensation,
and different cross-linking stages in acidic conditions, while in
basic conditions, it only shows as expected that all the processes
8
occur concomitantly. In addition, the application of a simple
kinetic model allows us to determine that the polycondensation
processes involved, even in acidic conditions, are certainly more
complicated than the classical ones usually described in polymer
chemistry, as has been suggested before.
1
as the main polycondensation controller. The electroactive
species irreversibly bind to the sol-gel colloids in the course
of their growing and thus the average mobility of the mobile
electroactive species formed in the course of the sol-gel
polycondensation reflects their average relative size and mass.
Usually the classical Cottrell law is verified, and the average
diffusion coefficient that can be extracted from the chrono-
amperograms is proportional to the hydrodynamic size of the
mobile species inside the sols or gels formed during all the
polycondensation course (as postulated by the classical Stokes-
Experimental Section
A. Preparation of Probe 1. (a) Preparation of ω-Alcenyl-
ferrocenes. Ferrocenemethanol is a commercial compound, but
was prepared from the aldehyde by a standard procedure, to
avoid excessive costs of appreciable quantities of this compound.
For the preparation of 4-hexenylferrocene, a typical preparation
is the following. In a flask under argon, 3 g of ferrocenemethanol
1
Einstein law, and we have demonstrated previously that the
-
2
3
recorded average current did differ only very little from the
average of the currents coming from all the species). It is true
that this simple relation may be inappropriate if the polymers
are highly concentrated, in which case a more complicated
model has to be considered. In certain situations, the fractal
dimension of the polymers also has to be taken into consider-
(1.36 × 10 mol) is dissolved in about 50 cm of freshly
distilled THF, and with a syringe is added 25 mL of a
-1
butyllithium solution (0.6 mol L in hexanes, 1 eq), and 5 min
later, is added 3 g of tosyl chloride at once under efficient
stirring. Meanwhile, in a separate flask has been prepared a
solution of pentenylmagnesium bromide by a conventional
procedure (from 3 g of pentenyl bromide, 1.5 eq), and this
Grignard reagent is then added to the former in-situ prepared
solution of methylferrocenetosylate; the flask was kept in ice
during the addition. After classical extraction, the resulting
ω-alcenylferrocene is purified on a silica gel column (eluent:
hexane).
3
ation. However, in sol-gel chemistry, often the polymers are
heavily cross-linked and the former analysis is applicable. We
present here an analysis of the sol-gel condensation of
tetramethoxysilane (TMOS), both in acidic and basic conditions.
Although this process has been thoroughly described in the past,
in several conditions, the techniques employed on this occasion
were often time- and equipment-demanding. Revisiting this test
reaction permits us to bring new light on it, as well as to
demonstrate the efficiency of the electrochemistry in the field
of polycondensation chemistry even more generally. In fact,
although the diffusion of small molecules in polymers has been
1H NMR (300 MHz, CDCl3) for ω-hexenylferrocene: δ )
1.22-1.48 ppm (m, 2H, Fc(CH2)2CH2), 1.48-1.60 ppm (m,
3
2H, FcCH2CH2), 2.09 ppm, (dd, J ) 7.0 Hz, 2H, Fc(CH2)3CH2),
3
2.34 ppm (t, J ) 7.0 Hz, 2H, FcCH2), 4.05-4.15 ppm (m,
9H, Fc), 4.95-5.07 ppm (m, 2H, Fc(CH2)4CHCH2), 5.84 ppm
5
2,3
thoroughly investigated, both in classical polymers and in sol
(ddd, J ) 7 Hz, Fc(CH2)4CHCH2).
6
13
gel polymers, the use of electrochemical methods to investigate
C NMR (75 MHz, CDCl3): δ ) 28.98 ppm, δ ) 29.54
7
transport in polymeric solutions remains rare, while monitoring
ppm, δ ) 30.68 ppm, δ ) 33.77 ppm, δ ) 67.18 ppm, δ )
68.22 ppm, δ ) 68.63 ppm, δ ) 89.55 ppm, δ ) 114.47 ppm,
δ ) 139.07 ppm.
polycondensation reactions is still an unexplored field to the
best of our knowledge.
1
The preparation of ω-butenylferrocene was identical, H NMR
*
Corresponding author.
(300 MHz, CDCl3): δ ) 2.25-2.40 ppm (m; 2H, FcCH2CH2),
1
0.1021/jp021872l CCC: $25.00 © 2003 American Chemical Society
Published on Web 01/16/2003

1322 J. Phys. Chem. B, Vol. 107, No. 6, 2003
Audebert et al.
SCHEME 1
a chronoamperometry are performed at this stage, so as to obtain
the diffusion coefficient D° of the probe, as well as to check
the electrochemical response. A solution of 0.2 mg of potassium
fluoride KF (3.4 µmol) in 0.5 mL of distilled water is then added
in the case of fluoride-catalyzed polycondensation, a solution
of 0.42 mg of (dimethylamino)pyridine (DMAP) (3.4 µmol) in
0
.5 mL of distilled water is then added, in the case of a base-
catalyzed polycondensation or 0.5 mL of pH 3.45 water (HCl-
acidified) in the case of an acid-catalyzed polycondensation. A
cyclic voltammetry and a chronoamperometry are then registered
every 2 min at the beginning and with more spaced time after,
to check the evolution of the electrochemical response (hence
the D value) of the systems. The experiments were made at the
laboratory temperature, which is controlled at 20 °C. We have
measured in a separate set of experiments the temperature
evolution, which is the following: The water addition is slightly
endothermic (3-4 °C, probably linked to the acetonitrile-water
mixing enthalpy), and then there is a slow rise, both due to the
weak exothermicity of the sol-gel reaction, and to the
equilibration at the room temperature. It takes about 10 min,
and is almost independent of the catalyst. Therefore, in the first
approximation, at such dilution conditions, the polycondensation
can be considered as athermic. Both experiments have been
repeated several times, and also checked against a sol-gel
preparation similar, but electroinactive (without probe added),
which showed that neither the probe nor the electrolyte salt
influenced drastically the gel behavior.
2
4
1
.40-2.55 ppm (m, 2H, FcCH2), 4.00-4.20 ppm (m, 9H, Fc),
.95-5.20 ppm (m, 2H, Fc(CH2)2CHCH2), 5.80-6.00 ppm (m,
H, Fc(CH2)2CHCH2).
1
3
C NMR (75 MHz, CDCl3): δ ) 28,83 ppm, δ ) 34.90
ppm, δ ) 66.93 ppm, δ ) 67.89 ppm, δ ) 68.34 ppm, δ )
8
8.64 ppm, δ ) 114.22 ppm, δ ) 138.30 ppm.
b) Preparation of 6-Triethoxysilyl-1hexyl-ferrocene (probe
). 6-Hexene-1-ferrocene (4.5 g, 16.8 mmol) and triethoxysilane
(
1
(6 mL; 31 mmol) are dissolved in freshly distilled toluene (20
mL) under argon and Sclenck conditions. Divinyltetramethyl-
siloxaneplatine (ABCR) dissolved in xylene (74 mL, 110 µmol)
is added. The solution is then refluxed for 14 h (110 °C).
The crude product is purified by chromatography on a silica
gel column after evaporation of the toluene (eluant: dichloro-
methane). A quantity of 3.5 g of 1 is obtained (48% yield)
(Scheme 1).
1H NMR (300 MHz, CDCl3) for 6-triethoxysilyl-1hexyl-
ferrocene: δ ) 0.62-0.68 ppm (m, 1H, CH2CH2Si(OEt)3), 1.24
ppm (t, 9H, J ) 7 Hz, Si(OCH2CH3)3), 1.30-1.55 ppm (m,
3
8
3
H, FcCH2(CH2)4CH2), 2.31-2.35 ppm (m, 2H, FcCH2(CH2)5);
Results and Discussion
3
.83 ppm (q, 6H, J ) 7 Hz, Si(OCH2CH3)3); 4.03-4.10 ppm
To functionalize the silicon alkoxides, the following ferro-
cenic probe 1 was prepared, according to Scheme 2.
(9H, m, Fc).
1
3
C NMR (75 MHz, CDCl3): δ ) 10.52 ppm, δ ) 18.41
ppm, δ ) 22.84 ppm, δ ) 29.35 ppm, δ ) 29.68 ppm, δ )
3
It should be noted that the reactive benzylic tosylate was not
isolated and was processed through a “one pot” reaction. The
overall yield of the probe was 20-25%, which is satisfactory
with regards to its further use, even on a scale of several grams.
The results of the electrochemical study are plotted as the
variation of the D/D° ratio, since: (1) We are looking at the
relative average variation of the grafted ferrocenic probes
diffusion coefficient through the reaction course (the pertinence
of the analysis in terms of average D has been previously
demonstrated in ref 1b, and (2) this allows one to get rid of the
experimental error introduced, for example, by the electrode
area. The chronoamperograms are given as the classical Cottrell
1.09 ppm, δ ) 33.12 ppm, δ ) 58.39 ppm, δ ) 67.05 ppm,
δ ) 68.13 ppm, δ ) 68.52 ppm, δ ) 89.63 ppm.
One should insist that great care has been taken to check the
purity of probe 1, since any trace of another ferrocene was likely
to make insignificant the electrochemical study. We believe that
the best way was to check that highly concentrated spots of the
probe on TLC did not let the slightest trace of another orange
spot (that turned green on oxidation) appear. This way is much
more sensitive than elementary analysis, which often, in
addition, is less precise in the case of siloxanes. In addition, no
partial hydrolysis of the compound was observed on the column,
which would have led to a loss at the purification stage. No
1
0
plots, where a straight line is obtained plotting the current i
1
29
1/2
SiOH group was observed on the H NMR spectra, and Si
NMR performed once showed the presence of only T0 (non
hydrolyzed) units. Lastly, one should take note that the choice
and the purity of the platinate catalyst is crucial, because the
use of other catalysts (from another company) resulted in only
hydrogenation of the alcene, instead of silylation.
Electrochemical Apparatus and Gel Preparation. The
electrochemical study of the sol gel polycondensation was
performed using a fast homemade potentiostat, equipped with
an ohmic drop compensation system, and connected to a
Tacussel pilot and a Nicolet digital oscilloscope, with a 1 mm
diameter platinum electrode. Probe 1 was further used for the
study of the probing of TMOS polycondensation. The following
protocol was adopted: TMOS (1 mL) and tetrabutylammonium
hexafluorophosphate (126 mg) are dissolved in 4 mL of
acetonitrile to obtain a 0.06 mol L electrolyte concentration.
The ferrocenic probe 1 (or ferrocene in the case of the free
probe) is then added (30 mg of 1 or 9.3 mg of ferrocene) to
against t , the slope being directly related to D, as detailed in
1
b
a previous paper . We have also calculated the real D° values
obtained for both probes used in the sol before the polymeri-
zation, although it should be kept in mind that the values are
obtained with a 20% precision, given the uncertainty on the
-5
2
-5
electrode surface. D° ) 1.7 × 10 cm /s for 1, and 3.7 × 10
2
cm /s for ferrocene; these values are in reasonable accordance
with the relative diameter of the two molecules, assuming that
the classical Stokes-Einstein law is followed. The cyclic
voltammograms are also registered in the course of the polym-
erization, and a typical feature is as in Figure 1, which shows
that the currents drastically diminish along the course of the
sol-gel polycondensation. This shows that all functionalized
probes attach themselves to the growing polymers, as it had
been previously shown in the case of optically active probes.11
Therefore the analysis in terms of diffusion coefficient analysis
is pertinent in our case.
9
-
1
Along the whole course of the polymerization reaction, (which
in the case of acid gels lasts over more than 1 day) one can
notice that small potential shifts may occur up to several tenths
-
2
-1
-2
reach a 1.4 × 10 mol L concentration in probe, and 10
mol L concentration in ferrocene. A cyclic voltammetry and
-
1

Polycondensation Kinetics of Silicon Alkoxides
J. Phys. Chem. B, Vol. 107, No. 6, 2003 1323
Figure 1. Cyclic voltammetry (V ) 5 V/s) and Cottrell plots obtained
by chronoamperometry during the polycondensation kinetics of silicon
alkoxides by acidic catalysis (HCl pH ) 3.45). Features represent the
evolution of the diffusion coefficient at (a) t ) 1 min 55 s, (b) t ) 21
h (which corresponds to gel time), (c) t over 8 days (room temperature,
2
0 °C). TMOS concentration was 1 mL in 4 mL of acetonitrile.
SCHEME 2
Figure 2. Representation of the relative D/D° variation with time for
a base-catalyzed gel, with experimental conditions as follows: potassium
fluoride (KF) (1/2000 vs TMOS) at room temperature. TMOS
concentration was 1 mL in 4 mL of acetonitrile. (a) Free ferrocene at
-
2
-2
1
0
M; (b) both free ferrocene at 10 M (symbols +, curve 1), and
-
2
probe 1 at 1.4 × 10 M (symbols • curve 2). Curve 1 comes from the
same data as Figure 1a, but is represented on a different scale in order
to allow straightforward comparison with the behavior of the func-
tionalized probe.
according to the slow evolution of its surface state. However,
the jump potential for the chronoamperometry experiments has
been chosen positive enough to get rid of possible problems
linked to this situation. Also the peak separation in the gel state
sometimes exceeds the 60 mV expected for a rapid system. This
may be due to slower charge transfer occurring with ferrocenes
taken inside large polymers; however, this has extremely little
influence on the chronoamperometry currents and does not
1
b
perturbate the analysis of the data.
Base-Catalyzed Sols and Gels. The results for a base-
catalyzed polycondensation are represented in Figure 1, which
shows that a drastic evolution of both the cyclic voltammograms
and the Cottrell plots takes place as the condensation goes on.
Figure 2 shows the dependence of the reduced diffusion
coefficient D/D°, in the case of fluoride gels. It is clear that the
polycondensation occurs on a long time scale, which extends
beyond after the gel time, and which in addition shows a
discontinuity at the gel point. The linearity of the middle part
of the graphs on a log-log scale is quite discernible, and both
of mVs (as in Figure 1). This is probably due to a shift of the
quasi-reference electrode, which is a silver wire, and whose
potential may be subjected to variations on a long time scale,

1324 J. Phys. Chem. B, Vol. 107, No. 6, 2003
Audebert et al.
graphs clearly feature three domains; the first domain is found
during the sol state (liquid), the second begins after gelation,
and the third domain starts after some aging of the gel.
In addition, Figure 2 displays, for comparative purposes, the
behavior of free ferrocene in exactly the same conditions. First
of all, it is clear from Figure 2a that only minor changes occur,
even noticeably after the gel point (as we and other groups
previously pointed out and studied.1,12 This shows as expected
that, in comparison to the changes related to the attachment of
the functionalized probe, the very minor changes in the
microviscosity of the liquid phase (sol, or included solution in
the case of the gels) plays no measurable role, as already verified
1
b
in a related work.
Sol Behavior. In the sol, no hindered diffusion exists, nor
does concentration depletion of the electroactive species occur.
Therefore the Stokes-Einstein law for the diffusion of the
relatively rigid sol-gel polymers applies. We have then: D ∼
-1
R , with R representing the mean hydrodynamic radius of the
polymer chains (radius of gyration), which varies with time.
We shall not forget that such rigid objects are made by either
diffusion-limited aggregation or reaction-limited aggregation,
and therefore are fractal. It is interesting to compare the
theoretical prediction for the variation of the average D value
with time, in the case of the two previously cited situations.
The diffusion-limited approach models the more generally
accepted forecasts that the objects formed have their fractal
dimension df equal to 1.78 or 1.98 (3), and a size variation
Figure 3. Same feature as Figure 2b, curve 2, for a DMAP-catalyzed
gel (DMAP 1/2000 vs TMOS).
point, where the hydrolysis may be expected to be completely
finished, a similar value is found.
It should be also pointed out that, given the average size of
the oligomers and polymers formed in the sols, it is very unlikely
that some redox probes became unable to transfer electrons,
due to the average diameter of the condensed species which is,
at the gel point, no more than about 10 times that of the
monomeric probe itself. This may perhaps change later,
especially in the aged gels.
Gels Behavior. After the gel point, a much sharper decrease
is noticed; however, it is very likely that after the infinite
network has formed, there are more and more immobile redox
species which no longer contribute to the electrochemical signal,
and therefore the decrease in the currents should be attributed
not only to the decrease of D, but also to a concentration
depletion of the electroactive moieties in solution. Further
experiments on ultramicroelectrodes are in preparation which
should help to clear up this point.
Acid-Catalyzed Sols and Gels. In the case of acidic catalysis,
the electrochemical response of the functionalized gels is quite
different. Two well-defined stages clearly appear, as can be seen
in Figure 4, with less defined intermediate stages.
During a first short period, D decreases very slowly, and then
slowly again up to the gel point which takes place at a stage
where the diffusion processes in the sol are still quite fast. Then
in the gel state, the D drop accentuates and becomes very fast
in the last aging step of the gel.
(1/df-1)
equal to R ∼ t
(for a 3-dimensional space). The variation
-
(1/df-1)
of D would come straightforward D ∼ t
. Therefore, the
D exponent variation should be 1.23 or 1.02, according to the
most satisfying model.3 This is clearly not the case, since a
slope of -0.65 ( 0.10 is observed. The reaction-limited models
take into consideration the time necessary for two objects to
react and aggregate together (which is obviously clearly related
,13
14
to polycondensation kinetics). According to Kolb and Jullien,
the mean mass (or size) of the objects formed should vary with
R
time according to M ∼ t , with R being a coefficient varying
continuously between 1.45 and 4.20. In fractal objects, the
1
/df
relation between the mass (size) and radius is R ∼ M with
the dimension df varying between 2.0 and 2.1 according to most
models. (One should also remark that a very simple approach
based on the classical Kirkwood’s model gives basically the
1/2
15
same result, since it forecasts a R ∼ M dependence; in this
case the fractal approach and the classical polymer approach
give the same result for a reasonable df ) 2 fractal dimension.
However, the high cross-linking degree of the polymers would
lead to favoring the fractal approach in the case of base-
-
R/df
catalyzed gels.) We arrive then at D ∼ t
. The case where
R is equal to 1.45 is considered closer to the diffusion-limited
process, while the limit of 4.20 is attained for very long reaction
times (and is also associated in the model with a longer and
longer induction period). According to our results, the base-
catalyzed gels behave in accordance with a fast reaction limited
process, since 1.45/2 ) 0.725, which is quite close to our 0.65
experimental value for the ln(D)-ln(t) slope on Figure 2.
Additional confirmation of this analysis comes also from the
fact that there is practically no induction period, in which case
the exponent of 1.45 is expected.
Sol Behavior. In the first induction period, which probably
corresponds to a phase where the hydrolysis of TMOS is still
important, a condensation phase is observable where D slowly
drops, the ln(D)/ln(t) slope being approximately -0.15. No
aggregation model could account for such a variation, but on
the other hand it is known that in acidic catalyzed polyconden-
4a,c
sation, essentially linear polymers are formed in a first stage,
after the hydrolysis is complete. The cross-linking processes
have been shown to occur in advanced stages of the synthesis.
We can then attempt to describe the sol state polymerization
by a classical polycondensation model.
One should notice that this is more obvious for the fluoride
catalysis, while the DMAP catalysis (Figure 3) shows a slightly
less linear behavior. This occurs possibly because in this case
the hydrolysis rate needs to be taken into greater consideration
in the overall polycondensation kinetics. However, if we
consider the tangency to the part of the curve before the gel
On the basis of the Stokes-Einstein law in the sol state, one
-
1
would again expect D ∝ R , R being the chain radius. Again,
for a polymer chain in solution in a good solvent, we have R ∝
(1/2) 15
n
, where n represents the number of monomer units, which

Polycondensation Kinetics of Silicon Alkoxides
J. Phys. Chem. B, Vol. 107, No. 6, 2003 1325
coefficient of the oligomers and polymers formed throughout
the polycondensation course in the sol allows us to propose an
aggregation process for the base-catalyzed reaction, although
in the case of the acid-catalyzed reaction a classical polycon-
densation mechanism gives a better account of the polymeri-
zation process. However, the change of regime observed in the
gel is harder to interpret, and the sharper drop of the experi-
mentally observed diffusion coefficient could also be attributed
to a concentration depletion due to immobilization of a part of
the redox probes. Further work is in progress to refine the first
results with the help of ultramicroelectrodes data, as well as to
investigate the polycondensation of hybrid gels.
References and Notes
(1) (a) Audebert, P.; Hapiot, P.; Sanchez, C.; Griesmar, P. J. Mater.
Chem. 1992, 2, 12. (b) Audebert, P.; Cattey, H.; Sanchez, C.; Hapiot, P. J.
Phys. Chem. 1998, 102, 1193.
(2) (a) Teraoka, I. In Polymer Solutions; J. Wiley: New York, 2002;
p 63. (b) de Gennes, P. G. In Scaling Concepts in Polymers Physics; Cornell
University Press: Ithaca, NY, 1979.
Figure 4. Same feature as Figure 2, for an acid-catalyzed gel. Same
experimental conditions as Figure 1 (HCl, pH ) 3.45).
(3) Meakin, P. In The Fractal Approach to Heterogeneous Chemistry;
Avnir, D., Ed.; J. Wiley and Sons, New York, 1989; pp 186-228.
(
4) (a) Devreux, F.; Boilot, J. P.; Chaput, F. Phys. ReV. A 1990, 41,
901. (b) Assink, R. A.; Kay, B. D. Annu. ReV. Mater. Sci. 1991, 21, 491.
c) Yang, H.; Ding, Z.; Xu, X. J. Non-Cryst. Solids 1989, 112, 449. (d)
6
(
therefore means D ∝ n-(1/2) . We can apply the classical result
(
1/2)
for a classical bifunctional polycondensation process, n ∝ t
Brunet, F.; Cabane, B. J. Non-Cryst. Solids 1993, 163, 211. (e) Fyfe, C.
A.; Aroca, P. P. Chem. Mater. 1995, 7, 1800.
16
(
at high n values) and so the expected dependence would be
D ∝ t . Given the very low dependence D ∝ t(0.15) for ln(D)
(1/4)
(5) (a) Long, J. W.; Velasquez, C. S.; Murray, R. W. J. Phys. Chem.
995, 100, 5492. (b) Pyati, R.; Murray, R. W. J. Am. Chem. Soc. 1996,
1
that we observed, it is likely that the polymer chains are much
more folded, and therefore that a R ∝ n( is much more likely,
showing that even in this case the chain solvatation is low.
However, it is also possible that the cross-linking process,
although not important, may affect the polycondensation kinetics
by slowing down the increase of the molecular weight.
Behavior in the Gels. As in the case of the base-catalyzed
gels, it is sure that the electrochemical response is quite sensitive
to the gelification, since the D curve displays a marked
discontinuity. More striking is the fact that the D dependence
drops very sharply during the aging stage of the gel. This
phenomenon is much steeper than in the base catalysis, and
therefore it is likely that restricted diffusion should occur after
the gelification, in addition to the concentration depletion.
1
18, 1743. (c) Velasquez, C. S.; Hutchinson, J. E.; Murray, R. W. J. Am.
1/3)
Chem. Soc. 1993, 115, 7896.
6) (a) Taussig, J. S.; Zambrano, P. J.; Wang, H.; Collinson, M. M.
Langmuir 1999, 15, 662. (b) Taussig, J. S.; Martin, S. A.; Collinson, M.
M. Chem. Mater. 1999, 11, 2594. (c) Audebert, P.; Griesmar, P.; Sanchez,
C. J. Mater. Chem. 1991, 1, 699.
(
(7) Williams, M. E.; Masui, H.; Long, J. W.; Malik, J.; Murray, R. W.
J. Am. Chem. Soc. 1997, 119, 1997.
(8) Wright, J. D.; Sommerdijk, A. J. M. In Sol-Gel Materials and
Applications; Gordon and Breach Science Publishers: Langhorne, PA, 2001;
p 20.
(9) Garreau, D.; Saveant, J. M. J. Electroanal. Chem. 1972, 35, 309.
(
10) Bard, A. J.; Faulkner, L. R. In Electrochemical methods, funda-
mentals and applications, 2nd ed.; J. Wiley: New York, 2001; pp 162-
63.
(11) (a) Sanchez, C.; Lebeau, B.; Maquet, J.; Toussaere, E.; Hierle, R.;
1
Zyss, J. J. Mater. Chem. 1994, 4, 1855; (b) Lebeau, B.; Sanchez, C.;
Brasselet, S.; Zyss, J. Chem. Mater. 1999, 9, 1012.
(
12) (a) Collinson, M. M.; Zambrano, P. J.; Wang, H.; Taussig, J. S.
Conclusion
Langmuir 1999, 15, 662; (b) Howells, A. R.; Zambrano, P. J.; Collinson,
M. M. Anal. Chem. 2000, 72, 5075.
We have shown for the first time that siloxane functionalized
ferrocenic probes were able to bind irreversibly to a polysiloxane
network during the polycondensation course. The electrochemi-
cal response of the functionalized ferrocene is therefore indica-
tive of the motion dynamics of the polymers formed in the
course of the reaction, and is in addition very sensitive to the
gelification. More detailed analysis of the average diffusion
(
(
13) Deutsch, J. M.; Meakin, P. J. Chem. Phys. 1983, 78, 2093.
14) (a) Jullien, R.; Kolb, M. J. Phys. A 1984, 17, L639; (b) Kolb, M.;
Jullien, R. J. Phys. Lett. 1984, 45, L977.
15) (a) Sperling, L. H. In Introduction to Physical Polymer Science; J.
(
Wiley: New York, 2001. (b) Sun, S. F. Physical Chemistry of Macromol-
ecules; J. Wiley: New York, 1994; pp 137-139.
(16) Mercier, J. P.; Marechal, E. In Chimie des Polym e` res; Presses
Polytechniques et Universitaires Romandes, EPFL Lausanne, 1993.