Organic-inorganic hybrid materials. Preparation and properties of dibenzo-18-crown-6 ether-bridged polysilsesquioxanes

Article abstract of DOI:10.1039/a909201c

Dibenzo-18-crown-6 ether derivatives bearing two (1, 2) or four (3) hydrolysable Si(OR)₃ groups have been synthesised. Their hydrolysis and polycondensation gave rise to new hybrid organic-inorganic materials incorporating dibenzo-18-crown-6 ether moieties covalently linked to silica by two or four Si-C bonds. The complexation of alkali metal cations (Na⁺ and/or K⁺) by these materials was investigated. A survey of the uptake of cations showed that four types of chelating sites exist within these materials, the ratio of which depends on the nature of the precursor (flexibility of the spacers between the benzene rings and the silicon centres and number of hydrolysable Si(OR)₃ groups) and also on the degree of condensation σ of the polysiloxane network. Furthermore, the complexation of K⁺ by the precursors 1, 2 and 3 and of Na⁺ by 1 were performed quantitatively. During the sol-gel polymerisation of these complexes it was shown that about 95% of the alkali cations were retained within the xerogel. This study proves that the two routes of incorporation of salts within these hybrid materials are not equivalent.

Full text of DOI:10.1039/a909201c

View Article Online / Journal Homepage / Table of Contents for this issue
J O U R N A L O F
Organic±inorganic hybrid materials. Preparation and properties of
dibenzo-18-crown-6 ether-bridged polysilsesquioxanes{
GeÂraud Dubois, Catherine ReyeÂ, Robert J. P. Corriu* and Claude Chuit
Laboratoire de Chimie Moleculaire et Organisation du Solide, UMR 5637 CNRS, UniversiteÂ
de Montpellier II, Sciences et Techniques du Languedoc, Place E. Bataillon, F-34095
Montpellier Cedex 5, France
C H E M I S T R Y
Received 22nd November 1999, Accepted 2nd March 2000
Dibenzo-18-crown-6 ether derivatives bearing two (1, 2) or four (3) hydrolysable Si(OR)3 groups have been
synthesised. Their hydrolysis and polycondensation gave rise to new hybrid organic±inorganic materials
incorporating dibenzo-18-crown-6 ether moieties covalently linked to silica by two or four Si±C bonds. The
complexation of alkali metal cations (Naz and/or Kz) by these materials was investigated. A survey of the
uptake of cations showed that four types of chelating sites exist within these materials, the ratio of which
depends on the nature of the precursor (¯exibility of the spacers between the benzene rings and the silicon
centres and number of hydrolysable Si(OR)3 groups) and also on the degree of condensation s of the
polysiloxane network. Furthermore, the complexation of Kz by the precursors 1, 2 and 3 and of Naz by 1
were performed quantitatively. During the sol±gel polymerisation of these complexes it was shown that about
95% of the alkali cations were retained within the xerogel. This study proves that the two routes of
incorporation of salts within these hybrid materials are not equivalent.
group), the monomer 2 having two hydrolysable Si(OEt)3
groups and the precursor 3 four hydrolysable Si(OMe)3 groups.
The hydrolysis and polycondensation of these monomers by
using the sol±gel process are reported in this paper. It is worth
noting that membranes incorporating dibenzo-18-crown 6-
ether moieties have been prepared by the sol±gel process to
facilitate the competitive transport of silver/copper ions.18 Our
purpose was different. While it was shown that macrocycles
such as 15-crown-5 ether and 18-crown-6 ether grafted to silica
have approximately the same af®nity for metal cations as the
corresponding unbound macrocycles in aqueous solutions,6,8 it
seemed to us interesting to investigate the binding ability of
dibenzo-18-crown-6 ether moieties incorporated within mate-
rials prepared by the sol±gel process towards Naz and/or Kz
ions in relation with the characteristics of the precursor:
number of hydrolysable Si(OR)3 groups, nature of the spacer
and degree of condensation of the network. It results from this
study that the rules of complexation within hybrid organic±
inorganic materials are different from those existing in
solution.
Introduction
The preparation of materials able to strongly complex metal
cations is of great interest1,2 as much from a fundamental point
of view (study of their reactivity and coordination chemistry
within the solid), as for the wide possibilities that open for
physical properties. They offer also potential applications to
perform selective metal cation separations. The strong and
selective interactions of crown ethers with speci®c alkali and
alkaline earth metal ions make these ligands ideal candidates
for such uses.2,3 Therefore their immobilization on organic or
inorganic polymers has been widely studied.4 In order to avoid
the gradual loss of the expensive ligand and the subsequent
decrease of the extraction effectiveness the covalent attachment
of the macrocyclic ligand to a solid support is required. For this
purpose, the binding of crown ethers to silica gel via pendant
SiOH groups has often been reported.5±10 However, this
method of attachment presents some drawbacks especially as
the content of organic moieties is often low and uncontrolled
and the dispersion in the ®nal materials is random.
Another method used to covalently attach a crown ether to
silica is the co-hydrolysis and co-polymerisation of a mixture of
a crown ether bearing one hydrolysable Si(OR)3 group with a
tetraalkoxysilane or an organotriethoxysilane11,12 by using the
sol±gel process.13 In this way, the concentration of chelating
groups in the solids is controlled but the dispersion of organic
moieties remains random.
Our interest in the study of organic±inorganic hybrid
materials14±16 led us to study the properties of solids obtained
by hydrolysis and polycondensation of crown ethers substi-
tuted by more than one hydrolysable Si(OR)3 group.
We have prepared as precursors the bissilylated dibenzo-18-
crown-6 ethers 1 and 2 and the tetrasilylated dibenzo-18-
crown-6 ether 3, the binding ability of the dibenzo-18-crown-6
ether being well known.17 1 presents a methylene group as
spacer between the benzene rings and two hydrolysable
Si(OiPr)3 groups while 2 and 3 present rigid spacers (ethenyl
Results
Bissilylated crown ether 119 was obtained by nickel-catalysed
cross-coupling reaction20 between triisopropyloxysilylmethyl-
magnesium chloride21 and the commercially available mixture
of 4,4'(5')-dibromodibenzo-18-crown-6 ether (Scheme 1). Bis-
silylated precursor 2 was prepared by Pd-catalysed cross-
coupling reaction of the mixture of 4,4'(5')-dibromodibenzo-
18-crown-6 ether with vinyltriethoxysilane (modi®ed Heck
reaction22) in 58% yield (Scheme 1). Tetrasilylated precursor 3
was prepared in the same way as 2 from the tetrabromodi-
benzo-18-crown-6 ether. The preparation of this last one by
bromination of dibenzo-18-crown-6 ether23 in CHCl3 provided
the tetrabromodibenzo-18-crown-6±hydrogen tribromide com-
plex24 as a side product in addition to the expected product. In
order to avoid this side reaction, acetic acid was used as solvent
to yield the expected tetrabromodibenzo-18-crown-6 ether in
80% yield. 3 was isolated in 35% yield as a very moisture
{Electronic supplementary information (ESI) available: elemental
analyses of xerogels. See http://www.rsc.org/suppdata/jm/a9/a909201c/
DOI: 10.1039/a909201c
J. Mater. Chem., 2000, 10, 1091±1098
This journal is # The Royal Society of Chemistry 2000
1091

View Article Online
Scheme 1
sensitive product from hexane extraction by using a Soxhlet
(Scheme 2).
reported in Table 2. The gel formation from 1 (compound
containing two hydrolysable Si(OiPr)3 groups) was conducted
with 10 mol% of HCl as catalyst while those of 2 (two
hydrolysable Si(OEt)3 groups) and 3 (four hydrolysable
Si(OMe)3 groups) were performed under nucleophilic condi-
tions (1 mol% of tetrabutylammonium ¯uoride) at different
temperatures. Geli®cation of 2 and 3 took place also in the
absence of catalyst (Table 2, entries 2 and 5). In all cases,
geli®cation occurred within 30 min except from 1 for which the
geli®cation appeared after several hours (v12 h) no doubt
because of the bulky Si(OiPr)3 groups.13b Dried gels obtained
after the work up given in the experimental section were
described by the following nomenclature: ®rst, an X to denote
xerogels followed by the number designation of the monomer
(1, 2 or 3), and last a letter (A, B or C) for the experimental
conditions of the reaction: A, without catalyst; B, 1% of TBAF
at room temperature; C, 1% of TBAF at 110 ³C in a sealed
tube.
The solid state 29Si NMR spectra of the xerogels displayed a
set of resonances lying between 259 ppm and 278 ppm
assigned to T1 [C±Si(OR)2OSi], T2 [C±Si(OR)(OSi)2], and T3
[C±Si(OSi)3] substructures (Table 2). The 29Si NMR spectra of
xerogels X2(A±C) are shown as examples in Fig. 1. The absence
of resonance corresponding to Q substructures26 (region of
2100 ppm) attested that no cleavage of Si±C bonds occurred
during the sol±gel process. The degree of condensation of the
hybrid materials, s, was evaluated by deconvolution of the
individual T resonances27,28 (Table 3). It appears that the
higher the temperature of reaction, the greater the degree of
condensation (compare in Table 3 entries 3, 4 and entries 6, 7).
It is also worth noting that the polycondensation of the
monomer 2 is complete in the presence of 1% of TBAF at
110 ³C. Elemental analysis showed that all the hybrid materials
deviated from the ideal stoichiometry based on totally
polycondensed silsesquioxane materials. Except for X2C
which is a very well condensed material, these analyses revealed
an excess of carbon and hydrogen (Table 3) indicating the
Complexation of Kz ions by the silylated macrocycles 1, 2
and 3 was performed in the presence of 1 equivalent of KSCN
in EtOH heated under re¯ux for 5 minutes to give quantita-
tively the corresponding 1 : 1 1±3[Kz] complexes. Treatment of
1 with 1 equivalent of NaSCN under the same conditions gave
quantitatively the corresponding 1 : 1 1[Naz] complex. The
formation of the complexes was ®rst indicated by a change of
solubility. While the compounds 1, 2 and 3 are rather soluble at
room temperature in pentane and hexane, the complexes are
1
not. All these complexes were characterised by their H NMR
spectra which displayed only one signal for the methylene
1
protons, except for 3[Kz], while the H NMR spectra of the
starting macrocycles showed 2 signals for the same protons
(Table 1). The complexation of Kz cations by the silylated
macrocycles 1, 2 and 3 as well as that of Naz by 1 was
con®rmed by diffuse re¯ectance IR spectroscopy (DRIFTS).
Indeed, the DRIFT spectra of complexed salts exhibited only
one absorption band for the CMN stretching vibration while the
DRIFT spectra for KSCN and NaSCN displayed two CMN
stretching vibrations and at lower frequencies (Table 1). It is
worth noting that the IR spectra of dibenzo-18-crown±6-
KSCN and NaSCN complexes25 displayed also one CMN
stretching vibration. A further indication for the formation of
the complex 1[Naz] was given from the solid state 23Na NMR
spectrum which displays a signal at d 225.6 while that of
NaSCN appears at d 210.6 under the same conditions. It is
also to be noted that the 23Na chemical shift for 1[Naz] is close
to that of dibenzo-18-crown-6±[Naz] complex (d 226.9).
Sol±gel processing of 1, 2 and 3 and characterisation of the
corresponding xerogels
The hydrolysis and polycondensation of precursors 1, 2 and 3
were performed in THF solutions (1 M in precursor) in the
presence of a stoichiometric amount of water. The experi-
mental conditions of geli®cation (catalyst, temperature) are
Scheme 2
1092
J. Mater. Chem., 2000, 10, 1091±1098

View Article Online
Table 1 1H NMR and DRIFT dataa for 1±3[Kz] and 1[Naz] complexes
1
1[Kz
]
1[Naz
]
2
2[Kz
]
3
3[Kz
]
1H NMR, d CH2O
DRIFT, nCN/cm21
3.96, 4.07
4.16
2051
4.11
2059
4.08, 4.20
4.18
2055
4.07, 4.24
4.10, 4.15
2056
anCN (cm21) for KSCN 2042, 2004; for NaSCN 2073, 2024; for dibenzo-18-C-6±[Kz] 2055; for dibenzo-18-C-6±[Naz] 2065.26
presence of residual hydroxy or alkoxy groups. The gap
between the experimental and the ideal values is higher
especially when the degree of condensation of the material is
low, as could be expected. Powder X-ray diffraction data for
xerogels showed that these materials are amorphous. The
surfaces areas determined by adsorption±desorption of N2
(BET)29 were found to be very low (v10 m2 g21) for all the
materials, whatever the experimental conditions of the
preparation.
for X1[Kz] as the powder X-ray diffraction pattern indicates
the presence of KCl.
The N2 BET surface areas of the materials incorporating
sodium and potassium salts were found to be very low
(v10 m2 g21) in all cases. Finally, elemental analysis revealed
that during the sol±gel polymerisation about 95% of alkali
cations were retained within the xerogel indicating that the
polycondensation occurred without distortion of the confor-
mation of the crown ether/salt complex. It is not surprising that
the strong interactions between the crown ether and the salt
survive the sol±gel process as it has been shown that weak
interactions remain unchanged during the sol±gel process.15(d)
Preparation and characterisation of the xerogels X1[Kz],
X1[Naz], X2[Kz] and X3[Kz]
Sol±gel polycondensations of complexes 1[Kz], 1[Naz], 2[Kz]
and 3[Kz] were performed at room temperature in 0.5 M THF
solution with a stoichiometric amount of water. While in the
presence of 10 mol% of HCl the gel formation of 1 took almost
12 h, those of 1[Kz] and 1[Naz] were very fast (v5 min). This
large difference in gel times could be attributed most probably
to a partial protonation of the free oxygen atoms of the crown
ether 1, thus decreasing the catalytic activity of HCl. Gelation
of 2[Kz] was conducted in the presence of 1% TBAF, while
that of 3[Kz] took place without any catalyst probably because
of the presence of four hydrolysable Si(OMe)3 groups per
crown ether instead of two hydrolysable Si(OEt)3 groups for
2[Kz]. Dried gels obtained after the work up given in the
experimental section are denoted by an initial X to denote
xerogel followed by the designation of the complex (Table 4).
The 29Si CP-MAS NMR spectra of the xerogels revealed that
there was no cleavage of the Si±C bond during the sol±gel
process as was observed for the free monomer and that the
condensation was incomplete (Table 4). DRIFT spectra of the
xerogels showed that the n(CMN) stretching band is very close to
that observed for the corresponding starting complexes in all
cases (compare n(CMN) in Table 1 for the complex and n(CMN) in
Table 4 for the corresponding xerogel). Another datum was
given for X1[Naz] from the solid state 23Na NMR spectrum
which displays one broad signal centred at d 226.3 (d 210.6 for
NaSCN under the same conditions) and another signal at d 0.1
attributed to NaCl. The presence of NaCl in the material was
con®rmed by powder X-ray diffraction data. Formation of
NaCl should be due to a displacement of complexed Naz ions
by the HCl used as catalyst for the gel formation, and should
then be limited to almost 10%. The same phenomenon occurs
Cation-binding properties of the xerogels
The binding ability towards sodium or potassium cations of
xerogels X1, X2(A±C) and X3(A±C) was investigated. Each
solid was treated with 1 equiv. of NaSCN or KSCN in EtOH
for 12 h at room temperature or heated under re¯ux. The
selectivity of the xerogels towards Naz and Kz ions was
likewise studied by competition experiments from a 1 : 1
mixture of NaSCN and KSCN in EtOH. The amount of
non-complexed cations was measured by ¯ame spectrophoto-
metry after ®ltration of the reaction mixture followed by
washing of the residue with EtOH until no more salt was
recuperated, and then evaporation of the solvent, and
dissolution of the recovered alkali metal salt in water. From
these data, the cations uptake into the xerogels was inferred
and reported in Tables 5, 6 and 7. Further con®rmation of the
incorporation of salts was given by elemental analyses. It is
important to note that the results given by these two methods
of titration are in good agreement. From DRIFT spectroscopy,
it was proved that the salts are not only incorporated within the
hybrid materials but complexed by the crown ether moieties,
only one CMN stretching band being observed (Tables 5, 6 and
7) at wavenumbers very close to those of the corresponding
complexes (Table 1). The complexation of the Naz ions was
also shown from the solid state 23Na NMR spectra of xerogels
incorporating NaSCN which display in all cases an up®eld
resonance (d lying between 219.2 and 226.6) (Tables 5, 6, and
7) with respect to the signal of NaSCN (d 210.6) under the
same conditions, the resonances being close to that observed
for the complex 1[Naz] (d 225.6).
The survey of cations uptake into X1 (Table 5) at room
temperature shows that:
(1) 72% (entry 2a) of crown ether moieties are able to
Table 2 Hydrolysis and polycondensation of precursors 1±3 (1 M in
THF) in the presence of a stoichiometric amount of water. 29Si CP-
MAS NMR data of the xerogels
29Si CP-MAS NMR (d)a
Entry Monomer Catalyst T/³C Xerogel T1
T2
T3
1
2
3
4
5
6
7
1
2
2
2
3
3
3
HCl 10%
Ð
TBAF 1% 22 X2B
TBAF 1% 110c X2C
22 X1 255
0 X2Ab 259
262
269
270
270
276
276
276
259
Ð
TBAF 1% 24 X3B
24 X3A
261
262
270
270
270
278
278
TBAF 1% 110c X3Cd 261
aMajor resonances in bold. b0.5 M THF solution of 2. cSealed tube.
d0.5 M THF solution of 3.
Fig. 1 29Si CP-MAS NMR spectra of xerogels X2A, X2B, and X2C.
J. Mater. Chem., 2000, 10, 1091±1098
1093

View Article Online
Table 3 Results of deconvolution analysis of the 29Si CP-MAS NMR spectra of xerogels and their elemental analyses
Elemental analysis
Entry
Xerogel
Degree of condensation s (%)
Ideal formula
22H26Si2O9
C24H26Si2O9
C24H26Si2O9
C24H26Si2O9
C28H28Si4O9
C28H28Si4O9
C28H28Si4O9
Experimental formula
1
2
3
4
5
6
7
X1
71
77
81
99
48
63
72
C
C24.3H31.6Si2O10.3
C26.6H31.6Si2O11.6
C25.9H30.2Si2O10.2
C23.7H26.4Si2O9.8
C33.8H44.8Si4O16
C32H42Si4O15.2
X2A
X2B
X2C
X3A
X3B
X3C
C32H39.7Si4O15.8
complex Kz ions
(2) 47% (entry 1a) of crown ether moieties are able to
complex Naz ions
These results are very different from those observed in
solution as in ethanol, the complexation of Kz ions on one
hand, and that of Naz ions on the other hand by the precursor
1 take place quantitatively to yield 1[Kz] and 1[Naz]
complexes.
Furthermore, about 12% of the crown ether moieties are able
to complex Naz in the presence of Kz (entry 3a). As about
47% of the crown ether moieties are able to complex Naz ions
in the absence of Kz ions (entry 1a), we can conclude that
about 35% of the complexing sites are able to complex both
ions but preferentially Kz ions. 78% of crown ethers being able
to complex Kz ions in the presence of Naz ions at room
temperature (entry 3a), we can infer that about 43% (78235) of
the crown ether moieties complex Kz ions. The results
obtained in ethanol heated under re¯ux are very similar
(Table 5, entries 1b, 2b, 3b). Thus, overall the results reported
in Table 5 suggest that within the material X1 crown ether
moieties exist with different binding properties:
(4) the percentage of B sites is always notably superior to the
percentage of C sites for a given xerogel
(5) the percentage of D sites seems mainly to depend on the
nature of the precursor
Discussion
To explain these experimental results we propose the following
interpretation.
The existence of the different types of chelating sites could be
due to the more or less important deformation undergone by
the crown ether moieties during the sol±gel process. One of the
causes for the deformations could very likely originate from the
existence of hydrogen bonds between the SiOH groups and the
oxygen atoms of the crown ethers and thus should be related to
the degree of condensation of the network.
The A sites which are able to complex both ions should be
relatively more ¯exible than the B and C sites. The latter should
be deformed in such a way that one type (B sites) complexes
Kz ions and the other (C sites) complexes Naz ions.
It can be noted that going from X2A to X2C or from X3A to
X3C, there is in both cases an increase in the percentage of A
sites, the increase being notably more important in the latter
case. As going from X2A to X2C and from X3A to X3C there is
an increase in the degree of condensation of the xerogels
(Table 3) with a concomitant decrease of the SiOH groups able
to bind the oxygen atoms of the crown ether, this con®rms that
there is a relation between the percentage of A sites and the
degree of condensation of the network. The greater the degree
of condensation s, the higher the percentage of A sites. The
hydrogen bonding should reduce the mobility of the crown
(1) the
A sites are able to complex both ions but
preferentially Kz ions (about 35%)
(2) the B sites complex Kz ions but not Naz ions (about
43%)
(3) the C sites complex Naz ions (about 12%) but not Kz
ions
(4) A fourth kind of binding site exits, i.e., the D sites, which
are not accessible for salts in EtOH or unable to chelate
Naz or Kz ions (about 10%).
An examination of the results reported in Tables 6 and 7
leads also to the evidence that these four types of chelating site
exist also in xerogels X2(A±C) and X3(A±C). An estimation of
the percentage of each type of site in a given xerogel is outlined
in Table 8.
ether and as
a consequence the possibility to change
conformation for the complexation of Naz or Kz ions. The
greater increase in A sites in the series X3(A±C) in comparison
with the series X2(A±C) agrees with this interpretation as for a
given degree of condensation s the number of SiOH groups in
the xerogels X3(A±C) is more important than in the xerogels
Further observations emerge from Table 8:
(1) the percentage of each type of chelating site depends on
the nature of the xerogels
X2(A±C) (the precursor
3 containing four hydrolysable
(2) the A sites are notably more important in the xerogel X1
than in the others
Si(OR)3 groups while the precursor 2 contains only two).
However, this explanation is not suf®cient to explain the
high percentage of A sites within X1, since, while the degrees of
condensation of X1 and of X2A are rather similar (Table 3), the
percentage of A sites within X1 (35%) is clearly more important
than that within X2A (7%). We consider that the high ratio of A
sites in X1 is due to the nature of the methylene spacer which is
(3) the percentage of A sites increases within a family of
xerogels [X2(A±B) or X3(A±C)] with the degree of
condensation s of the xerogels (reported in Table 3),
producing as a consequence a concomitant decrease of B
and/or C sites
Table 4 Hydrolysis and polycondensation of complexes. Spectroscopic data of the xerogels
29Si CP-MAS NMR dataa
T1 T2
Monomer
1[Kz
Catalyst
T/³C
Xerogel
X1[Kz
T3
DRIFT nCN/cm21
]
HCl 10%
HCl 10%
TBAF 1%
Ð
24
22
19
20
]
253
261.6
268
2056
2060
2057
2057
1[Naz
]
X1[Naz
]
262.6
270
269
2[Kz
3[Kz
]
]
X2[Kz
X3[Kz
]
]
279
278
261
aMajor resonances in bold.
1094
J. Mater. Chem., 2000, 10, 1091±1098

View Article Online
Table 5 Yield (%) of alkali metal cations incorporated within the xerogel X1 and spectroscopic data of the xerogel after reaction
Entry
Salt
T/³C
% Naz
% Kz
DRIFT nCN/cm21
23Na NMR d
1a
1b
2a
2b
3a
3b
NaSCN
KSCN
20
78
20
78
20
78
47,a 47b
43,a 46b
2068
2063
2059
2055
2054
2054
224.2
223.4
72,a 73b
77,a 76b
78,a 78b
74,a 76b
NaSCN±KSCN (1 : 1)c
NaSCN±KSCN (1 : 1)c
12a
10a
222
219.2
c
aMeasured by elemental analyses. bMeasured by ¯ame spectrophotometry. Competitive experiments.
more ¯exible than the ethenyl group bound to silicon in the
precursors 2 and 3. As the A sites involve a greater ¯exibility
than the other sites, this factor can explain the high ratio of A
sites in X1. Previous studies have shown the importance of
greater or lesser ¯exibility of the organic spacer in the
properties of hybrid materials.28
solids, are not equivalent. Thus, while the geli®cation of the
3[Kz] complex gives rise to a solid with 95% of its chelating
sites containing a Kz cation, the study of the binding ability of
xerogels X3(A±C) indicates that about 45% of crown ether sites
can complex neither Kz nor Naz ions. This reveals that the
geli®cation takes place with deformation of crown ether
moieties. This was con®rmed by competitive experiments
which have shown that each xerogel displays four types of
chelating site, the importance of which depends on the degree
of condensation s of the polysiloxane network, on the
¯exibility of the spacers between the benzene rings and the
silicon centres and on the number of hydrolysable Si(OR)3
groups borne by the precursor. Thus, it appears that the solid
alters the general rules of complexation of alkali metal cations
with crown ethers. In the solid, the nature of the precursor and
the degree of polycondensation of the network become the
dominating factors controlling the complexation. They are
both conducive to greater or lesser rigidity as well as to greater
or lesser deformation of the crown ether moieties in relation to
the formation of hydrogen bonds between the remaining SiOH
groups and the oxygen atoms of the crown ether.
The changes in the percentage of D sites with the nature of
the xerogel [about 10% for the xerogel X1, about 23% for the
xerogels X2(A±C) and about 43% for X3(A±C) (Table 8)] are
also of interest. First, it is worth noting that the lowest
percentage of D sites is observed within X1 which presents the
highest percentage of A sites, and the highest percentage of D
sites (48%) is within X3A, which presents the lowest percentage
of A sites (0%). Furthermore, as there are 23% of D sites in the
xerogel X2C, a very well condensed solid (s~99%), the
existence of the D sites cannot be explained by hydrogen
bonding between the oxygen atoms of the crown ether. We
consider that the percentage of D sites in a given xerogel
directly depends on the monomer which is essentially
characterised by the nature of the spacer (¯exible for 1, rigid
for 2 and 3) and by the number of hydrolysable Si(OR)3 groups
(two for 1 and 2 and four for 3). Thus, as the degrees of
condensation of X1 (s~71%) and of X2A (s~77%) are in the
same range, the low percentage of D sites (10%) within X1 in
comparison with X2A (23%) should be due to the ¯exible
methylene spacers of the monomer 1. The high percentage of D
sites within X3(A±C) should be essentially due to the presence
of four hydrolysable groups in the monomer 3. This gives rise
to a large number of remaining SiOH bonds in the xerogel and
as a consequence to a low ¯exibility of the organic moiety due
to the four SiO1.5 groups per chelating site. The case of xerogels
X2(A±C) is in between as they originate from the monomer 2
which presents a rigid spacer and contains only two hydro-
lysable Si(OR)3 groups.
Experimental
All reactions were carried out under argon by using a vacuum
line and Schlenk tube techniques. Solvents were dried and
distilled just before use. Melting points were determined with a
Gallenkamp apparatus and are uncorrected. IR data were
obtained on a Perkin-Elmer 1600 FTIR spectrophotometer by
the DRIFT method. The solution NMR spectra were recorded
on a Bruker AC-200 (29Si) and on a Bruker DPX-200 (1H and
13C). Chemical shifts (d) were referenced to Me4Si (1H, 13C,
29Si). Cross-polarization magic angle spinning (CP-MAS) 13C
and 29Si NMR spectra were recorded on a Bruker FTAM 300
operating at 75.47 MHz and 59.62 MHz respectively; CP-MAS
23Na NMR spectra were recorded on a Bruker DPX-200 at
79.39 MHz. In both cases, the repetition time was 10 s with a
contact time of 2 ms. FAB mass spectra [matrix, m-nitrobenzyl
alcohol (NBA)] were recorded on a JEOL JMS-D3000
spectrometer. Speci®c surface areas were determined by the
Brunauer±Emmett±Teller (BET) method on a Micromeritics
Gemini III 2375 analyser. Kz and Naz ions were titrated on a
Conclusion
We have shown that the xerogels prepared from the monomers
1, 2 and 3 gave rise to non-porous xerogels whatever the
experimental conditions of hydrolysis. It was observed that the
two routes of incorporation of salts, i.e. complexation of the
monomers followed by their geli®cation, or geli®cation of the
monomers followed by the incorporation of salts into these
Table 6 Yields (%) of alkali metal cations incorporated within the xerogels X2A, X2B and X2C at room temperature and spectroscopic data of the
xerogels after reaction
Entry
Xerogel
Salt
% Naz
31,a 31b
% Kz
DRIFT nCN/cm21
23Na NMR d
226.5
1a
1b
1c
2a
2b
2c
3a
3b
3c
X2A
NaSCN
KSCN
2060
2059
2060
2058
2058
2058
2058
2058
2058
52,a 52b
53,a 52b
NaSCN±KSCN (1 : 1)c
NaSCN
KSCN
24a
X2B
X2C
37,a 34b
225.8
222.8
55,a 49b
55,a 48b
NaSCN±KSCN (1 : 1)c
NaSCN
KSCN
26a
33,a 32b
56,a 55b
56,a 58b
NaSCN±KSCN (1 : 1)c
18a
c
aMeasured by elemental analyses. bMeasured by ¯ame spectrophotometry. Competitive experiments.
J. Mater. Chem., 2000, 10, 1091±1098
1095

View Article Online
Table 7 Yields (%) of alkali metal cations incorporated into the xerogels X3A, X3B and X3C at room temperature and spectroscopic data of the
xerogels after reaction
Entry
Xerogel
Salt
% Naz
12,a 10b
% Kz
DRIFT nCN/cm21
23Na NMR d
226.6
1a
1b
1c
2a
2b
2c
3a
3b
3c
X3A
NaSCN
KSCN
2063
2059
2059
2063
2058
2059
2063
2060
2058
37,a 39b
35,a 38b
NaSCN±KSCN (1 : 1)c
NaSCN
KSCN
15a
X3B
X3C
24,a 19b
224.6
223.2
45,a 45b
44,a 43b
NaSCN±KSCN (1 : 1)c
NaSCN
KSCN
17a
35,a 35b
48,a 48b
47,a 49b
NaSCN±KSCN (1 : 1)c
13a
c
aMeasured by elemental analyses. bMeasured by ¯ame spectrophotometry. Competitive experiments.
PHF 106 HYCEL apparatus. Elemental analyses were carried
out by the Service Central de Microanalyse du CNRS.
4,4'(5')-Bis(triisopropoxysilylmethyl)dibenzo-18-crown-6 ether±
NaSCN complex 1[Naz]
This complex was obtained in an identical manner to that
described above from 0.9 g (0.11 mmol) of 1 yielding a white
powder of 1[Naz] (0.99 g, 0.11 mmol, 100%). Mp 153±
154.5 ³C. 1H NMR (d, 250 MHz, CDCl3) 1.14 (d, 36H,
CH3), 2.07 (s, 4H, SiCH2), 4.11±4.13 [m, 22H, (OCH2 and
CH)], 6.68±6.75 (m, 6H, Ar); 13C NMR (d, 50 MHz, CDCl3,
{H}) 21.6 (CH2Si), 25.9 (CH3), 65.6 (CH), 67.1, 68.9 (OCH2),
110.8, 112.5, 121.7 (CH, Ar), 131.5, 132.9, 144.7, 146.8 (C, Ar);
29Si NMR (d, 40 MHz, CDCl3) 254.9 (s). CP MAS 23Na NMR
(d, 79.38 MHz): 225.6. IR (n/cm21 DRIFT, KCl) 2059. MS
(FABz, NBA) 819 [(MzNa)z, 100%]; 23 [(Naz), 42%]. Anal.
calc. for C41H68O12Si2NaSN: C, 56.10, H, 7.75, S, 3.65, N,
1.59%. Found: C, 56.09, H, 7.58, S, 3.49, N, 1.62%.
4,4',5,5'-Tetrabromodibenzo-18-crown-6 ether
In a 1000 ml three-necked ¯ask with a mechanical stirrer, an
isobar dropping funnel and a re¯ux condenser connected to a
gas absorption trap, 8 g (22 mmol) of dibenzo-18-crown-6 were
placed in an acetic acid solution (600 ml). An acetic acid
solution of bromide (100 ml, 98 mmol) was then rapidly added.
The solution became orange and release of HBr was observed.
The reaction mixture was heated under re¯ux at 95 ³C for 12 h.
The 4,4',5,5'-tetrabromodibenzo-18-crown-6 precipitated while
discoloration of the solution was progressively observed. When
the reaction mixture was at room temperature, the precipitate
was ®ltered off and washed three times with 50 ml of ether to
give 12 g (17.7 mmol, 80%) of 4,4',5,5'-tetrabromodibenzo-18-
crown-6 ether as a white powder. Mp 221±223 ³C. 1H NMR (d,
250 MHz, CDCl3) 3.98 (s, 8H, OCH2), 4.09 (s, 8H, OCH2), 7.04
(s, 4H, Ar); 13C NMR (d, 50 MHz, CDCl3, {H}) 70.1 (CH2),
70.8 (CH2), 116.3 (CH), 118.9 (CH), 149.7 (C). MS (IE, 70 eV):
m/z~676 ((Mz), 43%). Anal. calc. for C20H22Br4O6: C, 35.50,
H, 2.96, Br, 47.30%. Found: C, 35.75, H, 2.98, Br, 47.04%.
4,4'(5')-Bis(triethoxysilylethen-2-yl)dibenzo-18-crown-6 ether 2
5 g (9.65 mmol) of dibromodibenzo-18-crown-6 ether, 8.14 ml
(38.6 mmol) of vinyltriethoxysilane, 6.7 ml (48.3 mmol) of
triethylamine, 0.11 g (9.65 mol) of tetrakis(triphenylphosphine)
palladium and 80 ml of freshly distilled DMF were placed in a
150 ml Schlenk tube. The solution was heated under re¯ux for
48 h under argon. The solvent was then removed under vacuum
to give a brown oil which was dissolved in CCl4 (100 ml). After
®ltration of the solution and concentration of the ®ltrate, the
residue was again taken up in ether (2650 ml) and the solution
was again ®ltered. After removal of the solvent, a foam was
obtained which was washed with pentane (2615 ml). This
foam was pumped under vacuum for 12 h to give a beige
powder. The product was extracted from hot pentane
(2650 ml) to give after removal of the solvent 4.2 g (58%) of
2. Mp 69±71 ³C. 1H NMR (d, 250 MHz, CDCl3): 1.29 (t, 18H,
CH3), 3.91 (q, 12H, SiOCH2), 4.08 (m, 8H, OCH2), 4.20 (m,
8H, OCH2), 5.99 (d, 3JH,H~20 Hz, 2H, CHLCH), 6.81±7.04
4,4'(5')-Bis(triisopropoxysilylmethyl)dibenzo-18-crown-6 ether±
KSCN complex 1[Kz]
520 mg (0.65 mmol) of 1,20 63 mg (0.65 mmol) of KSCN and
10 ml of dry ethanol were placed in a 50 ml Schlenk tube and
heated under re¯ux for 5 min. The solvent was then removed
and the residue was washed twice with 20 ml of pentane to yield
580 mg of 1[Kz] (0.65 mmol, 100%) as a white powder. 1H
NMR (d, 250 MHz, CDCl3) 1.36 (d, 36H, CH3), 2.10 (s, SiCH2,
4H), 4.16±4.19 [m, 22H, (OCH2 and CH)], 6.71±6.78 (m, 6H,
Ar); 13C NMR (d, 50 MHz, CDCl3, {H}) 21.6 (CH2Si), 25.9
(CH3), 65.7 (CH), 67.1, 69.3 (OCH2), 110.8, 112.6, 121.9 (CH,
Ar), 131.7, 144.4, 146.6 (C, Ar); 29Si NMR (d, 40 MHz, CDCl3)
254.9 (s). IR (n/cm21, DRIFT, KCl) 2051. MS (FABz, NBA)
835 [(MzK)z, 40%]; 39 (Kz, 100%). Anal. Calc. for
C41H68O12Si2KSN: C, 55.08, H, 7.61, S, 3.58%. Found: C,
54.87, H, 7.23, S, 3.84%.
3
(m, 6H, Ar), 7.14 (d, JH,H~20 Hz, 2H, CHLCH); 13C NMR
(d, 50 MHz, CDCl3, {H}) 18.6 (CH3), 58.9 (SiOCH2), 69.1,
70.2 (OCH2), 111.5, 113.3, 115.5, 121.3, 131.7, 149.0, 149.9
(CHLCH, Ar); 29Si NMR (d, 40 MHz, CDCl3) 255.9 (s). Anal.
calc. for C36H56O12Si2: C, 58.69, H 7.61%. Found: C, 59.10, H,
7.73%.
Table 8 Estimation of the percentage of each type of site in xerogels after complexation of salts at room temperature
Xerogel
A sitesa
B sitesb
C sitesc
Accessible sites
D sitesd
X1
35
7
10
15
0
43
46
42
42
37
39
26
12
24
26
18
15
17
13
90
77
78
75
52
61
61
10
23
22
25
48
39
39
X2A
X2B
X2C
X3A
X3B
X3C
5
22
c
aAble to complex both ions but preferentially Kz
.
bComplex Kz ions. Complex Naz ions. dSites not accessible for salts or unable to chelate
Naz or Kz ions.
1096
J. Mater. Chem., 2000, 10, 1091±1098

View Article Online
4,4'(5')-Bis(triethoxysilylethen-2-yl)dibenzo-18-crown-6 ether±
for 5 days at room temperature, the solid was collected, and
washed with ethanol and ether three times. It was powdered
after each washing. The solid was dried under vacuum for 2 h
at 120 ³C yielding 1.84 g of a white powder. The experimental
conditions, the gel times and the 29Si NMR data of the xerogels
are reported in Table 2 and the elemental analyses in Table 3.
KSCN complex 2[Kz]
This complex was prepared in the same way as 1[Kz] starting
from 1.1 g (1.5 mmol) of 2 to yield a beige powder (1.25 g,
1.5 mmol, 100%). Mp 115 ³C (decomp.). 1H NMR (d,
250 MHz, CDCl3) 1.28 (t, 18H, CH3), 3.87 (q, 12H,
SiOCH2), 4.18 (m, 16H, OCH2), 6.01 (d, 3JH,H~20 Hz, 2H,
Preparation of xerogels X1[Kz], X1[Naz], X2[Kz] and
X3[Kz]. The xerogels X1[Kz], X1[Naz], X2[Kz] and
X3[Kz] were prepared according to the same procedure. The
IR and CP MAS 29Si NMR data are indicated in Table 4.
3
CHLCH), 6.80±7.07 (m, 6H, Ar), 7.23 (d, JH,H~24 Hz, 2H,
CHLCH); 13C NMR (d, 50 MHz, CDCl3, {H}) 18.7 (CH3),
59.0 (SiOCH2), 67.5, 67.7, 69.2 (OCH2), 109.0, 111.3, 116.5,
121.3, 131.9, 147.3, 147.8, 148.6 (CHLCH, Ar); 29Si NMR (d,
40 MHz, CDCl3) 256.3 (s). IR (n/cm21 DRIFT, KCl) 2055.
MS (FABz, NBA) 775 [(MzK)z, 25%], 39 [(Kz), 100%].
Anal. calc. for C37H56O12Si2KSN: C, 53.30, H, 6.72%. Found:
C, 53.49, H, 6.83%.
Complexation of Kz and Naz by the xerogels. In all cases, a
given xerogel was stirred with an EtOH solution of NaSCN, or
of KSCN or of both of them at room temperature or at re¯ux
for 12 h. The quantity of salts introduced into the xerogel was
calculated for a completely condensed material. After 12 h of
reaction, the mixture was treated according to the following
procedure which is given as an example. The IR and CP MAS
23Na NMR data for xerogels X1 after treatment are indicated
in Table 5. The corresponding data for X2(A±C) are in Table 6
and for X3(A±C) in Table 7.
4,4',5,5'-Tetrakis(trimethoxysilylethen-2-yl)dibenzo-18-crown-6
ether 3
7 g (10.4 mmol) of tetrabromodibenzo-18-crown-6 ether,
12.64 ml (41.4 mmol) of vinyltrimethoxysilane, 15 ml
(0.1 mol) of triethylamine, 0.12 g (0.1 mmol) of tetrakis(tri-
phenylphosphine)palladium and 150 ml of freshly distilled
DMF were placed in a 500 ml Schlenk tube. The solution was
heated under re¯ux for 60 h under argon. The solvent was then
removed under vacuum to give a brown oil. This oil was taken
up again in CCl4 (150 ml) and the solution was ®ltered. After
®ltration and concentration of the ®ltrate, the residue was
again taken up in ether (2650 ml) and the solution was ®ltered.
After removal of the solvent, a foam was obtained which was
pumped under vacuum for 12 h to give a brown powder. The
product was extracted for 24 h from hot hexane (200 ml) by
using a soxhlet. After removal of the solvent 3.3 g (3.5 mmol,
34%) of 3 were obtained. Mp 70±72 ³C. 1H NMR (d, 250 MHz,
CDCl3) 3.65 (s, 36H, CH3), 4.07 (m, 8H, OCH2), 4.24 (m, 8H,
OCH2), 5.88 (d, 3JH,H~19 Hz, 4H, CHLCH), 6.94 (s, 4H, Ar),
Complexation of Naz by the xerogel X1. 61 mg of the
xerogel X1, 9.9 ml of an EtOH solution of NaSCN (solution
1.2661022 M) and 10.1 ml of EtOH were placed in a ¯ask. The
mixture was stirred at room temperature for 12 h and was then
®ltered. The solid was washed four times with 10 ml of EtOH. It
was then dried at 120 ³C for 12 h under 20 mmHg to give 57 mg
of a white powder.
29Si NMR (d) 261, 270. 23Na NMR (d) 224.2. IR (n/cm21
DRIFT, KCl) 2068. Anal. calc. for C23H26Si2O9NaSN: C,
48.33, Na, 4.03, Si, 9.81%. Found: C, 50.89, Na, 1.90, Si,
9.85%.
3
7.46 (d, JH,H~19 Hz, 4H, CHLCH); 13C NMR (d, 50 MHz,
References
CDCl3, {H}) 51.0 (CH3), 68.4, 70.3 (OCH2), 111.1 (CH, Ar),
117.9 (CHLCH), 130.4 (C, Ar), 146.9 (CHLCH); 149.8 (C, Ar).
29Si NMR (d, 40 MHz, CDCl3) 253.6 (s). Anal. calc. for
C40H64O18Si4: C, 50.85, H, 6.78%. Found: C, 51.95, H, 6.96%.
1
2
3
I. O. Sutherland, Chem. Soc. Rev., 1986, 15, 63.
J. M. Lehn, Science, 1985, 227, 849.
R. M. Izatt, J. S. Bradshaw, S. A. Nielsen, J. D. Lamb,
J. J. Christensen and D. Sen, Chem. Rev., 1985, 85, 271.
M. Takagi and H. Nakamura, J. Coord. Chem., 1986, 15, 53.
T. G. Waddell and D. E. Leyden, J. Org. Chem., 1981, 46, 2406.
J. S. Bradshaw, R. L. Bruening, K. E. Krakowiak, B. J. Tarbet,
M. L. Bruening, R. M. Izatt and J. J. Christensen, J. Chem. Soc.,
Chem. Commun., 1988, 812.
4
5
6
4,4',5,5'-Tetrakis(trimethoxysilylethen-2-yl)dibenzo-18-crown-6
ether±KSCN complex 3[Kz]
This complex was prepared in the same way as 1[Kz] starting
from 1.41 g (1.5 mmol) of 3 to give a beige powder (1.55 g,
1.49 mmol, 99%) of 3[Kz]. 1H NMR (d, 400 MHz, CDCl3) 3.57
(s, 36H, CH3), 4.10 (m, 8H, OCH2), 4.15 (m, 8H, OCH2), 5.85
(d, 3JH,H~19 Hz, 4H, CHLCH), 6.88 (s, Ar, 4H), 7.45 (d,
3JH,H~19 Hz, 4H, CHLCH). 13C NMR (d, 100 MHz, CDCl3,
{H}) 49.6 (CH3), 66.3, 67.7 (OCH2), 107.5 (CH, Ar), 117.5
(CHLCH), 129.0 (C, Ar), 145.1 (CHLCH), 146.3 (C, Ar). 29Si
NMR (d, 40 MHz, CDCl3) 254.1 (s). IR (n/cm21 DRIFT,
KCl) 2056. MS (FABz, NBA) 983 [(MzK)z, 20], 39 [(Kz),
100]. MS (FAB2, NBA) 58 [(SCN)2, 100%]. Anal. calc. for
C41H64O18Si4KSN: C, 47.26, H, 6.15, S, 3.07, N, 1.34%.
Found: C, 48.51, H, 5.90, S 4.82, N, 0.98%.
7
8
9
C. W. McDaniel, J. S. Bradshaw, K. E. Krakowiak, R. M. Izatt,
P. B. Savage, B. J. Tarbet and R. L. Bruening, J. Heterocycl.
Chem., 1989, 26, 413.
J. S. Bradshaw, K. E. Krakowiak, B. J. Tarbet, R. L. Bruening,
J. F. Biernat, M. Bochenska, R. M. Izatt and J. J. Christensen,
Pure Appl. Chem., 1989, 61, 1619.
R. M. Izatt, R. L. Bruening, B. J. Tarbet, L. D. Grif®n,
R. L. Bruening, K. E. Krakowiak and J. S. Bradshaw, Pure Appl.
Chem., 1990, 62, 1115.
10 T. Okada and T. Usui, J. Chem. Soc., Faraday Trans., 1996, 92,
4977.
11 P. Lacan, C. Guizard, P. LeGall, D. Wettling and L. Cot,
J. Membr. Sci., 1995, 100, 99.
12 K. Kimara, T. Sunagawa and M. Yokoyama, Chem. Commun.,
1996, 745.
13 (a) C. J. Brinker and G. W. Sherer, Sol±gel Science, Academic
Press, London, 1990.; (b) L. L. Hench and J. K. West, Chem. Rev.,
1990, 90, 33.
14 D. A. Loy and K. J. Shea, Chem. Rev., 1995, 95, 1431.
15 (a) R. J. P. Corriu and D. Leclercq, Angew. Chem., Int. Ed. Engl.,
1996, 35, 1421; (b) R. Corriu, Polyhedron, 1998, 17, 925;
(c) R. Corriu, C. R. Acad. Sci. SeÂr. IIc, 1998, 83; (d) G. Cerveau
and R. J. P. Corriu, Coord. Chem. Rev., 1998, 178, 1051.
16 G. Cerveau, R. J. P. Corriu and C. Lepeytre, Chem. Mater., 1997,
9, 2561.
Sol±gel processing.
General procedure for the preparation of xerogels X1, X2(A±
C), X3(A±C). All the geli®cations were carried out from a THF
solution of the monomer. A stoichiometric amount of water
was ®rst added followed by the required amount of the catalyst.
The procedure for the xerogel X2A is given as an example. To a
THF solution (8.14 ml) of 2 (3 g) placed in a 31 ml ¯ask were
added 0.22 ml of water followed by 0.4 ml of a 0.1 M THF
solution of TBAF. The ¯ask was cooled at 0 ³C and maintained
at this temperature until gel formation (v30 mm). After ageing
17 J. C. Pedersen, J. Am. Chem. Soc., 1967, 89, 7017.
18 M. Barboiu, C. Luca, C. Guizard, N. Hovnanian, L. Cot and
G. Popescu, J. Membr. Sci., 1997, 129, 197.
J. Mater. Chem., 2000, 10, 1091±1098
1097

View Article Online
19 C. Chuit, R. J. P. Corriu, G. Dubois and C. ReyeÂ, Chem.
Commun., 1999, 723.
24 E. Shchori and J. Jagur-Grodzinski, J. Am. Chem. Soc., 1972, 94,
7957.
20 R. J. P. Corriu and J. P. Masse, J. Chem. Soc., Chem. Commun,
1972, 144; K. Tamao, K. Sumitani and M. Kumada, J. Am. Chem.
Soc., 1972, 94, 4374.
21 D. J. Brondani, R. J. P. Corriu, S. El Ayoubi, J. J. E. Moreau and
M. Wong Chi Man, J. Organomet. Chem., 1993, 451, C1.
22 H. Yamashita, B. L. Roan and M. Tanaka, Chem. Lett., 1990, 12,
2175.
25 S. N. Poonia, J. Am. Chem. Soc., 1974, 96, 1012.
26 M. Magi, E. Lippmaa, A. Samoson, G. Engelhardt and
A. R. Grimmer, J. Phys. Chem., 1984, 88, 1518.
27 K. J. Shea, D. A. Loy and O. Webster, J. Am. Chem. Soc., 1992,
114, 6700.
28 G. Cerveau, R. J. P. Corriu, C. Lepeytre and P. H. Mutin,
J. Mater. Chem., 1998, 8, 2707.
23 K. H. Pannell, W. Yee, G. S. Lewandos and D. C. Hambrick,
J. Am. Chem. Soc., 1977, 99, 1457.
29 S. Brunauer, P. H. Emmet and E. Teller, J. Am. Chem. Soc., 1938,
60, 309.
1098
J. Mater. Chem., 2000, 10, 1091±1098