Para-Nitroaniline-functionalized chromophoric organic-inorganic hybrid materials

Article abstract of DOI:10.1039/c2jm15407b

An advanced procedure for the one-pot synthesis of organic-inorganic hybrid materials via combination of sol-gel process and nucleophilic aromatic substitution reaction (S_NAr) of 4-fluoronitrobenzene and 3-aminopropyltrimethoxysilane is described. With this advanced procedure both S_NAr-reaction and sol-gel process can be accomplished in the same reaction vessel due to the sol-gel precursor tetraethoxysilane (TEOS) acting as solvent during the first reaction step. Via extensive NMR spectroscopic studies it is proven for the first time that - contrary to common belief - hydrogen fluoride (HF), which is formed as a by-product in the S_NAr-reaction, is not trapped by any bases present but is rather trapped by both of the silane species and serves as a catalyst during the subsequent sol-gel process. The chromophoric system of the resulting xerogels is well protected against aggressive chemicals, i.e. strong acids, by the silica matrix, which predestines these materials for pigment applications. Given that a high chromophore content is highly desirable in these applications, we show that the chromophore content of the final xerogel can be varied by modification of the organosilicon precursor:TEOS ratio or by using trialkoxy-silanes bearing two or three amino functions, whereby the latter option is more favourable. Monodisperse core-shell particles with identical chromophore content but consisting of a pure silica-core and a p-nitroaniline functionalized shell with a diameter of about 200 nm can also be prepared using this advanced procedure.

Full text of DOI:10.1039/c2jm15407b

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links <
C
Journal of
Materials Chemistry
Cite this: J. Mater. Chem., 2012, 22, 3839
www.rsc.org/materials
PAPER
para-Nitroaniline-functionalized chromophoric organic–inorganic hybrid
materials†
Katharina Ladewig,ab Andreas Seifert,a Harald Hahn,a Michael Hietschold,c Norbert Moszner,d
d
Peter Burtscher and Stefan Spange
a
*
Received 24th October 2011, Accepted 16th December 2011
DOI: 10.1039/c2jm15407b
An advanced procedure for the one-pot synthesis of organic–inorganic hybrid materials via
combination of sol–gel process and nucleophilic aromatic substitution reaction (SNAr) of
4-fluoronitrobenzene and 3-aminopropyltrimethoxysilane is described. With this advanced procedure
both SNAr-reaction and sol–gel process can be accomplished in the same reaction vessel due to the sol–
gel precursor tetraethoxysilane (TEOS) acting as solvent during the first reaction step. Via extensive
NMR spectroscopic studies it is proven for the first time that—contrary to common belief—hydrogen
fluoride (HF), which is formed as a by-product in the SNAr-reaction, is not trapped by any bases
present but is rather trapped by both of the silane species and serves as a catalyst during the subsequent
sol–gel process. The chromophoric system of the resulting xerogels is well protected against aggressive
chemicals, i.e. strong acids, by the silica matrix, which predestines these materials for pigment
applications. Given that a high chromophore content is highly desirable in these applications, we show
that the chromophore content of the final xerogel can be varied by modification of the organosilicon
precursor : TEOS ratio or by using trialkoxy-silanes bearing two or three amino functions, whereby the
latter option is more favourable. Monodisperse core–shell particles with identical chromophore content
but consisting of a pure silica-core and a p-nitroaniline functionalized shell with a diameter of about 200
nm can also be prepared using this advanced procedure.
available by dissolution of the desired dye in one of the
Introduction
components required for the sol–gel process prior to initiation of
gelation. However, this purely physical/sterical interaction
between the dye and matrix means that encapsulated dyes can be
extracted by solvents or can be attacked by contaminants causing
loss and/or degradation of the chromophore with time. There-
fore, most applications utilize type II hybrid materials, simply
because the organic component (dye) is protected from external
influences such as organic solvents and/or other aggressive
compounds. For the preparation of such materials functional
alkyl alkoxysilanes with covalently tethered chromophores are
required. The synthesis of those compounds is laborious, time
consuming, and expensive.1–17
Coloured organic–inorganic hybrid materials find application in
many fields, but most particularly in a variety of optical and
sensing applications or in the pigment industry.1–17 Generally,
they are comprised of organic chromophores, i.e., organic dyes,
and an inorganic, structure- and property-determining matrix.
According to Sanchez’s classification of hybrid materials,5 there
are two modalities by which an organic dye can be attached to an
inorganic matrix, via encapsulation (type I) or covalent linkages
(type II). Hybrid materials of the former type are readily
aInstitute of Chemistry, Chemnitz University of Technology, Strasse der
Nationen 62, D-09111 Chemnitz, Germany. E-mail: stefan.spange@
chemie.tu-chemnitz.de
bDepartment of Chemical and Biomolecular Engineering, The University of
Melbourne, Parkville, VIC, 3010, Australia. E-mail: katharina.ladewig@
unimelb.edu.au
Recently, we described an alternative experimental procedure
for the synthesis of such class II hybrid materials, which
combines the nucleophilic aromatic substitution (SNAr) reaction
of fluoro-aromatic compounds and aminofunctionalized tri-
alkoxysilanes with a consecutively occurring sol–gel process in
the presence of tetraethoxysilane (TEOS) and in some cases
a base in a very convenient one-pot synthesis.18–21 The reaction
concept is illustrated in Scheme 1. The SNAr reaction of fluoro-
aromatic hydrocarbons such as 4-fluoronitrobenzene (4-FNB)
with primary or secondary aminotrialkoxysilanes results in the
formation of a chromophoric push–pull system if the substituent
cInstitute of Physics, Chemnitz University of Technology, Strasse der
Nationen 62, D-09111 Chemnitz, Germany
dIvoclar Vivadent AG, Bendererstrasse 2, FL-9494 Schaan, Liechtenstein
† Electronic supplementary information (ESI) available: Eqn (S1):
mathematical equation for the calculation of 3; Fig. S1: UV/Vis spectra
of heat-treated xerogels; Fig. S2: plot of lmax of solubilized
p-nitroaniline functionalized xerogels vs. pH; Fig. S3 and S4:
13C-{1H}-CP-MAS NMR spectra of DAS- and TAS-derived xerogels.
See DOI: 10.1039/c2jm15407b
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 3839–3852 | 3839

View Article Online
Scheme 1 Synthetic procedure for the preparation of chromophoric xerogels. The yellow coloured N-(3-trimethoxysilylpropyl)-para-nitroaniline
chromophore is formed through the nucleophilic aromatic substitution (SNAr) reaction of 4-fluoronitrobenzene (4-FNB) and 3-amino-
propyltrimethoxysilane (APS) in tetraethoxysilane (TEOS) as solvent using an external base B, i.e., N,N-dimethyl-n-octylamine (step I). The final
product is obtained after initiation of the sol–gel process with water and controlled aging of the gel (step II), drying and purification by Soxhlet
extraction (step III).
in para- or ortho-position to the fluorine atom is strongly electron
withdrawing.22
was synthesized from DMOA and hydrofluoric acid and
vacuum-dried over phosphorus pentoxide prior to use.
Since TEOS acts as a solvent in this first step, no additional
solvent (i.e., alcohol) is required for the SNAr reaction and the
consecutive sol–gel process can be accomplished in the same
reaction vessel. In recent papers we demonstrated that this
concept works well for the immobilization of a broad variety of
molecular structures of chromophores.18–21 However, some open
questions remained, namely (i) the role of fluoride during the
SNAr reaction, (ii) the extent to which the intermediate
product—the chromophore—is incorporated into the final
product, (iii) whether the xerogel syntheses are suitably repro-
ducible, (iv) the xerogels are stable to thermal and/or other
external stimuli, and (v) whether it is possible to increase the dye
content and if so how. The HF formed during the SNAr reaction
is thought to serve as catalyst during the sol–gel process, hence
may not need to be removed or trapped by a base (e.g., tertiary
amine). However, its role during this first step of the synthesis
process, the SNAr reaction, and its catalyzing effect on the sol–
gel process in this particular system are not fully understood and
need to be elucidated further. As shown in previous publications,
the colour and other physicochemical properties of the produced
materials can be adjusted in a wide range by variation of the
fluoro-compound and/or the synthesis conditions.18–21 However,
their application as pigments also calls for highly stable xerogels,
which are able to withstand a variety of chemical and mechanical
stresses, while at the same time maximum colour intensity needs
to be realisable in blends using minimal quantities of these
materials. In this manuscript we systematically study the role of
HF, the thermal and pH-stability of as-synthesized xerogels as
well as ways to achieve particle monodispersity and the different
means by which the colour intensity can be increased.
Synthesis of p-nitroaniline functionalized xerogels
One-pot synthesis method. N-(3-Trimethoxysilylpropyl)-para-
nitroaniline (pNA) functionalized xerogel syntheses were per-
formed following the two different pathways described in
Scheme 3. Isolation of the product was achieved by rotary
evaporation at 70 ꢀC and 15 mbar, followed by further dew-
atering using a rotary furnace (7 h at 70 ꢀC and 0.1 mbar),
Soxhlet extraction with acetone (64 h) and finally drying of the
product until constant weight in a vacuum drying cabinet at 40
ꢀC. DAS- and TAS-derived pNA-functionalized xerogels were
synthesized according to pathway B (Scheme 3).
Core–shell particles. 1.5 L ethanol, 120 mL ammonia (25%), 7.1
mL water and 74.4 mL (367.9 mmol) TEOS are stirred for five
hours at room temperature to give the silica core of the final
€
product (Stober particles). 5.19 g (36.8 mmol) 4-FNB, 7.71 g (37.0
mmol) ꢀTEOS and 6.85 g (38.2 mmol) APS are stirred for 5.5 hours
at 150 C under inert argon atmosphere and mixed with 20 mL
pure ethanol after cooling down to room temperature. The yellow
€
solution is added to the Stober particles under vigorous stirring.
Ammonia and ethanol are removed under reduced pressure after
16 hours of stirring and a yellow coloured powder is obtained.
Further processing was performed as for the other xerogels.
Synthesis of M1 ([N-propyl]-4-nitroaniline)
2.26 g (16.0 mmol) 4-FNB was added to 20 mL of cold (ꢁ18 ꢀC)
n-propylamine. After 10 minutes 2.4 mL (17.2 mmol) trime-
thylmethoxysilane was added and the mixture was heated to 60
ꢀC for one hour. The solvents were removed under vacuum at
110 ꢀC to give the pure product in 95% yield with a melting point
Experimental section
1
of 62 ꢀC. H NMR (250 MHz, d, ppm, CDCl3) ¼ 8.08 (2H, m),
Materials
3
6.51 (2H, m), 4.5 (1H, s, broad), 3.18 (2H, J(1H–1H) ¼ 7.1 Hz,
3J(1H–1HNH) ¼ 5.6 Hz), 1.69 (2H, tq, 3J(1H–1H) ¼ 7.4 Hz,
3J(1H–1H) ¼ 7.1 Hz), 1.02 (3H, t, 3J(1H–1H) ¼ 7.4 Hz). 13C-{1H}
NMR (62.5 MHz, d, ppm, CDCl3) ¼ 153.4, 137.8, 126.5, 110.9,
45.1, 22.4, 11.5.
APS (97%, Lancaster), TEOS (98%, Acros Organics), N,N-
dimethyl-n-octylamine (DMOA) and pyridine were distilled
under reduced pressure and stored in dry argon atmosphere prior
to use. N-Propylamine (99%, Acros Organics), 3-(2-amino-
ethylamino)propyltrimethoxysilane (DAS, >98%, Fluka), (3-
trimethoxysilylpropyl)diethylenetriamine (TAS, 95%, ABCR),
trimethylmethoxysilane (95%, ABCR), hexamethyldisiloxane
(>98%, Alfa Aesar), N-propylethylenediamine (99%, Acros
Organics) and diethylenetriamine (>98%, Merck) were used as
received. N,N-Dimethyl-n-octylammonium fluoride (DMOAF)
Synthesis of M2 ([N1,N2-bis(4-nitrophenyl)-N1-propylethane-
1,2-diamine])
3.18 g (22.4 mmol) 4-FNB, 0.58 g (5.6 mmol) N-propylethyle-
nediamine and 1.44 g (13.8 mmol) trimethylmethoxysilane were
3840 | J. Mater. Chem., 2012, 22, 3839–3852
This journal is ª The Royal Society of Chemistry 2012

View Article Online
refluxed at 160 ꢀC for 34 h. Ethanol was added to the mixture and
the yellow-brown coloured solid was filtered off and again boiled
with ethanol and filtered off. Yellow needles with a melting point
of 176–178 ꢀC in a yield of 25% were obtained. 1H NMR
(250 MHz, d, ppm, d6-DMSO) ¼ 8.00 (4H, m), 7.34 (1H,
t, 3J(1H–1H) ¼ 6.0 Hz, NH), 6.79 (2H, m), 6.66 (2H, m), 3.65 (2H,
imaging, gold was evaporated onto the samples to render them
electrically conductive. Transmission electron microscopy
(TEM) was performed using a Philips CM 20 FEG microscope
with the samples being fixed on carbon grids.
Thermogravimetry. TG measurements were performed in air
between 40 and 900 ꢀC at a heating rate of 10 K minꢁ1 using
a Thermogravimetric Analyser 7 (Perkin Elmer).
3
t, J(1H–1H) ¼ 6.5 Hz), 3.39 (4H, m), 1.56 (2H, m), 0.88 (3H, t,
3J(1H–1H) ¼ 7.3 Hz). 13C-{1H} NMR (62.5 MHz, d, ppm, d6-
DMSO) ¼ 154.6, 153.0, 136.2, 135.8, 126.5, 126.3, 110.9, 52.5,
49.4, 39.9, 20.0, 11.2.
Specific surface area. Nitrogen adsorption isotherms were
measured at 77.3 K using a Sorptomatic 1900. Specific surface
areas were calculated using the Brunauer–Emmett–Teller (BET)
approximation. Single point measurements were performed on
Synthesis of M3 ([N1,N2,N3-tris(4-nitrophenyl)-
diethylenetriamine])
€
an Areameter II setup (Sohnlein Instruments).
1.36 g (13.2 mmol) diethylenetriamine, 5.57 g (39.5 mmol) 4-FNB
and 2 mL hexamethyldisiloxaneꢀwere refluxed for 5 h at 140 ꢀC,
ꢀ
Results and discussion
20 h at 150 C and 10 h at 190 C. The crude product was dis-
solved in DMSO, precipitated by addition of ethanol and filtered
off. The procedure was repeated six times to obtain a light brown
powder with a melting and decomposition temperature of 269–
270 ꢀC in 36% yield. 1H NMR (250 MHz, d, ppm, d6-DMSO) ¼
8.05 (2H, m), 7.95 (4H, m), 7.34 (2H, t, 3J(1H–1H) ¼ 5.8 Hz), 6.85
(2H, m), 6.62 (4H, m), 3.66 (4H, t, 3J(1H–1H) ¼ 6.4 Hz), 3.42 (4H,
SNAr reaction of 4-FNB and aminofunctionalized
trialkoxysilanes
The use of TEOS as both solvent during the SNAr reaction of
fluoro-aromatic compounds with aminosilanes and precursor for
the subsequently induced sol–gel process is the key feature of our
previously described one-pot synthesis method.18–21 Tetraalkox-
ysilanes are solvents of moderate polarity whereas nucleophilic
aromatic substitutions of amines with fluoro-aromatic
compounds succeed in polar solvents such as dimethyl sulfoxide
or pyridine.26–28 Selective combinations of solvent mixtures and
additional bases were studied in order to elucidate the influence
of both polarity and basicity of the environment on the conver-
sion of APS with 4-FNB. The progress of individual reactions
3
dt (3J(1H–1H) ¼ 6.4 Hz); J(1H–1H) ¼ 5.8 Hz). 13C-{1H} NMR
(62.5 MHz, d, ppm, d6-DMSO) ¼ 154.5, 152.9, 136.3, 136.2,
126.5, 126.3, 111.2, 49.9, 39.7.
Characterization
NMR spectroscopic measurements. Solid state NMR spectra
were collected at 9.4 T on a Bruker Avance 400 spectrometer
equipped with double-tuned probes capable of MAS (magic
angle spinning). 13C-{1H}-CP-MAS NMR spectra were
measured at 100.6 MHz in 4 mm standard zirconium oxide
rotors (BRUKER) spinning at 12 kHz. Cross-polarization with
a contact time of 5 ms was used to enhance the sensitivity. The
recycle delay was 4 s. Spectra were referenced externally to tet-
ramethylsilane (TMS) as well as to adamantane as secondary
standard (38.48 ppm for 13C).23 29Si-{1H}-CP-MAS NMR spec-
troscopy was performed at 79.5 MHz using 7 mm rotors spinning
at 5 kHz. The contact time was 3 ms and the recycle delay 3 s.
Shifts were referenced externally to tetramethylsilane (0 ppm)
with the secondary standard being tetrakis(trimethylsilyl)silane
1
was observed by H NMR spectroscopy.
Firstly, pure chromophoric N-(3-trimethoxysilylpropyl)-para-
nitroaniline (pNA) was synthesized via the reaction of 4-FNB
ꢀ
and APS at 100 C in freshly distilled pyridine as described by
Claude et al.29 Complete conversion was achieved after 4 h. As
shown in Fig. 1 multiple signal sets are found in the 1H and 13C-
{1H} NMR spectra for both the methoxy groups of the
(CH3O)3Si moiety and the methylene group adjacent to the
silicon atom.
This result was wholly reproducible even when the reactions
were performed under absolutely water-free and inert conditions.
Additionally, the number of hydrogen signals observed for the
anticipated (CH3O)3Si moiety is less than nine. These observa-
tions indicate that a substitution reaction occurs at the silicon
atom. Theoretically, one molecule of HF is liberated for each
pNA molecule formed. As suggested previously,30 this can lead to
fluorine substitution at the silicon atom of both the product and
starting reagent APS (Scheme 2).
1
(ꢁ9.8, ꢁ135.2 ppm). All spectra were collected with H decou-
1
pling using a TPPM pulse sequence. H decoupled quantitative
29Si-{1H}-MAS NMR spectra were collected with 70ꢀ rf pulses
and recycle delays of 300 s.24 Decomposition of individual signals
was performed using DMFit software provided by Massiot
et al.25
According to Tacke et al. the substitution reaction occurs via
a zwitterionic, pentavalent silicon species,31–34 and methanol
formed by either APS or pNA is readily detectable using various
suitable methods such as gas chromatography or 1H NMR
spectroscopy, a fact not considered by Claude et al.29 Further, we
observed that the equimolar reaction between 4-FNB and APS in
the substance without pyridine takes place quantitatively. This
indicates that the HF formed is not trapped by APS via an acid–
base reaction that forms ammoniumpropyltrimethoxysilane
fluoride, because if this is the case the reaction would only
UV/Vis spectroscopic measurements. UV/Vis absorption
spectra were collected using a UV/Vis MCS 400 diode-array
spectrometer (Carl Zeiss, Jena) connected to a diffuse reflectance
accessory cell via glass-fiber optics. Solutions were measured by
the transmission technique using quartz cuvettes with an optical
pathway length of 2 mm.
Electron microscopy. Scanning electron micrographs (SEMs)
were recorded on a Philips SEM 515 instrument. Prior to
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 3839–3852 | 3841

View Article Online
Scheme 2 SNAr-reaction of 4-FNB with APS forming N-(3-trimethox-
ysilylpropyl)-para-nitroaniline (pNA) in pyridine as solvent, and possible
subsequent reactions of pNA (route i) and APS (route ii) with hydrogen
fluoride that is formed during the SNAr-reaction.
150 ꢀC being necessary before 90% conversion of the reactants is
achieved. Hence, we investigated the influence of a variety of
experimental parameters, e.g., molar ratios, temperatures, and
additional bases, on the reaction rate. Namely pyridine and the
stronger base N,N-dimethyl-n-octylamine (DMOA) were used in
approximately equimolar concentrations to 4-FNB to accelerate
the conversion. Time conversion plots for the various reaction
parameters studied are displayed in Fig. 2. From the plots it can
be concluded that the addition of external bases to the reaction
mixture is not really necessary (compare runs 1, 2 and 3), since
the increase in reaction rate is almost outweighed by the slowing
effect the lowered educt concentrations (due to the addition of
the base) have on the reaction rate. Surprisingly pyridine was
observed to have a considerably stronger effect on the reaction
Fig. 1 (A) 1H and (B) 13C-{1H} NMR spectra and peak assignment for
the anticipated product, N-(3-trimethoxysilylpropyl)-para-nitroaniline
(pNA), of the SNAr-reaction between 4-fluoronitrobenzene (4-FNB) and
3-aminopropyltrimethoxysilane (APS) [reaction in pyridine at 100 ꢀC,
spectra measured in d6-acetone (#)].
proceed to 50% since the remaining APS would be fully
protonated at this point and the formed ammonium hydrogen
fluoride is not capable of SNAr reactions. In short, the formed
HF cannot be trapped by any bases for the reason that acid–base
reactions are equilibrium reactions while the formation of fluoro-
silanes is not. It can therefore be concluded that any HF that is
formed as the reaction progresses does indeed react further with
any of the alkoxy silane species present in the reaction mixture
(routes (i) and (ii) in Scheme 2).
Secondly, the reaction of APS and 4-FNB was carried out
using TEOS as the solvent. Changing the solvent from pure
pyridine to TEOS slows down the reaction rate considerably,
with reaction times of 20 h or more and temperatures as high as
Fig. 2 Progress of the SNAr-reaction of 4-FNB with APS as a function
of time under various reaction conditions [TEOS : tetraethoxysilane,
DMOA : N,N-dimethyl-n-octylamine, ratios given refer to molar ratios
of the compounds; lines are a guide for the eye of the reader only].
3842 | J. Mater. Chem., 2012, 22, 3839–3852
This journal is ª The Royal Society of Chemistry 2012

View Article Online
rate than DMOA, even though it is a weaker base (compare runs 2
and 3). Having a ten-fold excess of TEOS in the reaction mixture
slows down the reaction rate so significantly that 90% conversion
of the reagents is only achieved after more than 20 h of refluxing
reactions the formed HF can undergo with the alkoxysilanes,
they also undergo substitution reactions amongst themselves,
probably catalyzed by the presence of the fluoride ions.
Si–C bond cleavage as observed by Moreau et al.35 for the SN2-
reaction of chloromethyltrimethoxysilane with amines was not
observed. After reaction of a 1 : 1 mixture of APS and 4-FNB for 5
hat150ꢀCtogivethepNAproductandsubsequentinitiationofthe
sol–gel process only T-signals and no Q-signals were found in the
29Si-{1H}-CP-MAS NMR spectra of the xerogel which corrobo-
rates our conclusion that all Si–C bonds survived the reaction.
ꢀ
the reaction mixture at 150 C (runs 4 and 5). In contrast, the
reaction of 4-FNB and APS in the substance (without TEOS) at
130 ꢀC is completed within three hours (run 6).
However, omitting the TEOS completely renders the reaction
mixture quite difficult to handle due to the formation of subli-
mates and causes inconsistent results. Thus, we decided on an
equimolar ratio of 4-FNB, APS, and TEOS for the first reaction
step, the SNAr-reaction. After complete conversion of the reac-
tants is achieved with certainty, i.e., after five hours at 150 ꢀC, the
remaining nine equivalents of TEOS are added to the mixture
and the sol–gel process is initiated. A 29Si-{1H} NMR spectrum
of the reaction mixture just prior to the addition of the remaining
nine equivalents of TEOS reveals many different silicon species
originating from the variety of substitution reactions that HF
can undergo with TEOS and APS. We were able to assign all
observed signals to their corresponding molecular structure
Effect of HF on gelation times
Fluoride ions are known to be very effective sol–gel catalysts
since they strongly influence both the hydrolysis and the
condensation reactions of alkoxysilanes.36–39 The effect of the
fluoride concentration on the gelation time was studied using
a model system consisting of APS and TEOS (in a molar ratio of
1 : 10) and the model fluoride N,N-dimethyl-n-octylammonium
fluoride (DMOAF). The model fluoride was dissolved in ethanol
(equivalent to three times the amount of water used to start
the sol–gel process; nEtOH ¼ 3nwater) and subsequently added to
the TEOS–APS mixture. Gelation times were measured from the
moment water (equimolar to all hydrolysable Si–OR bonds;
nwater ¼ 3nAPS + 4nTEOS) was added to the mixtures containing
various amounts of DMOAF until the initially clear solutions
appeared cloudy. As shown in Table 2 gelation occurred after
approximately 15–18 min in the reference experiment without
DMOAF (run 1), whilst concentrations of fluoride below 1 ppm
were already able to halve this time to about 8 min (run 4). A
further increase of the fluoride concentration resulted in gelation
times as short as one second (run 7). However, it is undesirable
that the gelation occurs too fast since this can result in inho-
mogeneities in the final xerogel. One simple means of prolonging
gelation times is to increase the ethanol amount used for the sol–
gel process (runs 6, 6a, and 6b). Based on the discussed experi-
ments an ethanol to water ratio of 10 : 1 was deemed most
appropriate for the xerogel syntheses.
1
element by H–29Si NMR correlation experiments (Table 1).
The occurrence of various signals attributable to mixed alkoxy
silane species clearly indicated that apart from the substitution
Table 1 Silicon species found in the reaction mixture of 4-FNB, APS,
and TEOS after five hours at 130 ꢀC as determined by liquid 29Si-{1H}
NMR spectroscopy in d6-acetone
Silicon speciesa
29Si chemical shift
Coupling constants
R–Si(OMe)3
ꢁ42.6 ppm
ꢁ43.8 ppm
ꢁ45.0 ppm
ꢁ46.2 ppm
ꢁ46.8 ppm
ꢁ47.8 ppm
ꢁ48.9 ppm
ꢁ78.5 ppm
ꢁ79.4 ppm
ꢁ80.3 ppm
ꢁ81.2 ppm
ꢁ82.1 ppm
Not observed
ꢁ86.8 ppm
ꢁ87.6 ppm
Not observed
R–Si(OMe)2(OEt)
R–Si(OMe)(OEt)2
R–Si(OEt)3
R–SiF(OMe)2
R–SiF(OMe)(OEt)
R–SiF(OEt)2
Si(OMe)4
Si(OMe)3(OEt)
Si(OMe)2(OEt)2
Si(OMe)(OEt)3
Si(OEt)4
1J(29Si–19F) ¼ 272.2 Hz
1J(29Si–19F) ¼ 272.4 Hz
1J(29Si–19F) ¼ 272.6 Hz
SiF(OMe)3
SiF(OMe)2(OEt)
SiF(OMe)(OEt)2
SiF(OEt)3
1J(29Si–19F) ¼ 195.4 Hz
1J(29Si–19F) ¼ 196.3 Hz
Synthesis and characterization of p-nitroaniline functionalized
xerogels
Having derived the optimum conditions for both SNAr-reaction
and sol–gel process the reproducibility of xerogel syntheses was
a
R ¼ (CH2)3NHC6H4(p-NO2).
Table 2 Gelation times of 1 : 10 molar mixtures of APS and TEOS after addition of water as a function of fluoride concentration (DMOAF) in ethanol
as solvent. Water was used in an equimolar ratio to the number of Si–OR groups present in the mixture
#
Mole fraction of fluoridea/ppm
Molar ratio EtOH : waterb
Molar ratio silicon : fluoride
Gel time
1
2
3
4
5
6
6a
6b
7
0
0.008
0.04
0.08
24
107
55
25
1130
3 : 1
3 : 1
3 : 1
3 : 1
3 : 1
3 : 1
7 : 1
17 : 1
3 : 1
1 : 0
7 000 000 : 1
140 000 : 1
70 000 : 1
2500 : 1
15–18 min
12–14 min
11–12 min
8–8.5 min
50 s
30 s
80 s
23 min
1 s
560 : 1
560 : 1
560 : 1
55 : 1
a
b
Mole fraction of fluoride (DMOAF) ¼ nDMOAF/(nDMOAF + nAPS + nTEOS + nwater + nEtOH). nwater ¼ 3nAPS + 4nTEOS
.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 3839–3852 | 3843

View Article Online
tested next. Scheme 3 explains the two variations of the synthesis
procedure used for this purpose. Despite the benefits an excess of
APS can have on the conversion rate of the SNAr-reaction
equimolar ratios of 4-FNB and APS were employed during all
xerogel syntheses since the primary amino-groups of non-reacted
APS can have a negative effect in subsequent applications of
these xerogels, e.g., if they are to be used as fillers in reactive
polymer systems cured via radical polymerization. Therefore,
removal of any 4-FNB remaining after xerogel synthesis via
Soxhlet-extraction is preferred, because it occurs quantitatively.
By means of the two procedures shown in Scheme 3 a series of
p-nitroaniline-functionalized xerogels was synthesized. The
simplest way to verify the reproducibility of xerogel syntheses is
to weigh the final product and determine the yield of the reaction.
However, the catch here is that the exact molecular weight of the
final xerogel is unknown. The ‘‘idealized’’ structure of the
product as depicted in Scheme 1 has a molecular weight of 832.15
g molꢁ1, but this value is based on the assumption that all Si–OR
bonds are hydrolyzed and all Si–OH groups are condensed
during the reaction, and the xerogel does not contain any
impurities such as residual solvent or water. The graph displayed
in Fig. 3 illustrates the relationship between the mass of xerogel
obtained and the feed amount of 4-FNB during our experiments.
The linear correlation of these two values indicates very good
reproducibility of the synthetic process. The slope of the linear
regression was determined to be 850.09 g molꢁ1, which represents
the ‘‘real’’ molecular weight of the formed xerogels. This slightly
larger value than the ‘‘ideal’’ molecular weight indicates that
either the sol–gel process or the SNAr-reaction does not proceed
to complete conversion or there may be a loss of the organic
chromophore during the long extraction procedure with acetone.
Quite possibly a combination of all three of the above mentioned
reasons causes the observed difference between ‘‘ideal’’ and
‘‘real’’ molecular weight of the xerogels.
Solid state 29Si-{1H}-CP-MAS NMR measurements revealed
the presence of Q2- and Q3-signals in the spectra of all xerogels,
which clearly indicates the incomplete hydrolysis and conden-
sation of the ethoxy groups of TEOS (Fig. 4A). This type of
sensitivity enhanced NMR experiment uses a cross-polarization
1
(CP) technique to transfer the magnetization from H to 29Si
nuclei; hence signals of silicon species which are in close prox-
imity to protons are particularly amplified. As a result T-,
Q2- and Q3-signals are overestimated, while Q4-signals which
originate from completely condensed silicon species are under-
estimated. In contrast, quantitative solid state 29Si-{1H}-MAS
NMR spectra reflect the exact composition of the xerogel more
Scheme 3 Illustration of the different synthesis pathways employed.
3844 | J. Mater. Chem., 2012, 22, 3839–3852
This journal is ª The Royal Society of Chemistry 2012

View Article Online
the error of the method due to the signal to noise ratio of the
spectrum). Further, the sole presence of T3-, Q3-, and Q4-signals
with T2- and Q2-signal intensities being negligible shows that the
siloxane network of the xerogel is highly condensed. The
observed Q3/Q4 signal intensity ratio is 1 : 2.4, which means that
more than 70% of the ethoxy groups of tetraethoxysilane were
converted to ^Si–O–Si^ bridges. The nearly complete absence
of T2-signals indicates that virtually all methoxy groups of 3-
aminopropyltrimethoxysilane are completely hydrolyzed and
condensed.
These conclusions are confirmed by the presence of 13C NMR
signals for non-hydrolyzed ethoxy groups of TEOS in the solid
state 13C-{1H}-CP-MAS NMR spectra of the xerogels (Fig. 5).
13C-{1H}-CP-MAS NMR spectra of all xerogels were measured
to confirm the formation of the desired p-nitroaniline-function-
alized chromophore and its covalent attachment to the xerogel
matrix. As exemplified in Fig. 5, all bar two 13C NMR signals
that are observed originate from the pNA-functionalized chro-
mophore. The two additional signals at 60 and 18 ppm (Fig. 5,
signals 5 and 8, respectively) can be assigned to the non-hydro-
lyzed ethoxy groups of TEOS. No 13C NMR signals commonly
assigned to non-hydrolyzed CH3O-groups of APS were detected,
and neither were any 13C NMR signals of residual 4-fluoroni-
trobenzene, which indicates high conversion rates and/or high
efficiency of the purification procedure.
Fig. 3 Mass of xerogel obtained as a function of the feed amount of 4-
FNB. The linear correlation verifies synthesis reproducibility, while the
slope can be used to determine the apparent molecular weight of para-
nitroaniline (pNA) functionalized xerogels [n ¼ 18 syntheses according to
procedures A (4 syntheses) and B (14 syntheses) as shown in Scheme 3].
accurately, but at the cost of longer measuring times. As shown
in Fig. 4B the signals for the T2- as well as the Q2-silicon species
all but vanished in the quantitative spectrum since their mole
fractions are far below 0.5 mol%. Furthermore, the non-sensi-
tivity enhanced 29Si-{1H}-MAS NMR spectrum showed a weak
signal at about ꢁ175 ppm which was not observable in the 29Si-
{1H}-CP-MAS NMR spectrum. According to the literature this
signal can be assigned to a hexacoordinated silicon-fluoro-
species.40 Deconvolution of the quantitative 29Si-{1H}-MAS
NMR spectrum shown in Fig. 4B was performed with DMFit
software provided by Massiot et al.25 Particularly noteworthy
amongst the results of this analysis, which are summarized in
Table 3, is that the calculated ratio of T-signal to the sum of the
Q-signals amounts to 1 : 10.8. This value accurately reflects the
APS : TEOS ratio of 1 : 10 used during synthesis (with respect to
Quantitative determination of the chromophore content is
achieved by dissolving a defined amount of dry xerogel in 1 M
sodium hydroxide (NaOH) solution. When dissolved in 1 M
NaOH the yellow coloured p-nitroaniline derivative shows
a strong UV/Vis absorption at 416 nm. In the case of xerogels it is
more practical to determine their mass-related extinction coeffi-
cient or mass-related chromophore content, F, instead of their
molar extinction coefficient, 3, since their real molecular weight
differs from their nominal molecular weight (eqn (1); refer to eqn
(S1) in the ESI† for details on how to calculate 3):
ꢀ
ꢁ
EV
md
3
L
F ¼
¼
;
(1)
M
g cm
The value F is directly proportional to the chromophore content
in a solid xerogel and specific for each individual batch. It is only
useful to compare the optical properties of xerogels containing
identical chromophores as different chromophores may be
influenced by the dissolution process in different ways. Table 4
summarizes the F-values that were determined for the various
synthesized xerogels. The reproducibility of the chromophore
content is very good as indicated by the low standard deviations.
It should be noted that for the samples produced without using
TEOS as the solvent during the SNAr-reaction slightly lower F-
values were found. This can be explained by small amounts of the
yellow pigment precursor sublimating at higher temperature as
observed for the reaction of 4-FNB and APS in the substance,
which we discussed earlier. In order to determine the maximum
F-value, Fmax, for p-nitroaniline functionalized xerogels with
a chromophore–TEOS ratio of 1 : 10 the model compound N-
propyl-4-nitroaniline (M1) was synthesized. Based on the molar
extinction coefficient of this purely organic chromophore (3 ¼
21 270 L molꢁ1 cmꢁ1) and the ‘‘ideal’’ molecular weight of 1 : 10
Fig. 4 Solid state 29Si-{1H} NMR spectra of a xerogel synthesized
according to path B of Scheme 3. (A) Sensitivity enhanced 29Si-{1H}-CP-
MAS NMR [7068 scans, contact time 3 ms, relaxation delay 3 s,
measuring time 6 h]. (B) Quantitative 29Si-{1H}-MAS NMR [549 scans,
relaxation delay 300 s, measuring time 2 days]. The signal labelled with
the hash sign (#) originates from hexacoordinated silicon species.
xerogels (832.12 g molꢁ1), Fmax can be as high as 25 L gꢁ1 cmꢁ1
.
The discrepancies between Fmax and experimental F-values
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 3839–3852 | 3845

View Article Online
Table 3 Quantitative analysis of the solid state 29Si-{1H}-MAS NMR spectrum shown in Fig. 4B by the structure of silicon speciesa
29Si NMR signal
T2
T3
Q2
Q3
Q4
Structure
Chemical shift/ppm
Percentage found
ꢁ59.0
ꢁ66.0
ꢁ93.0
ꢁ103.0
ꢁ112.0
n.d.
8.5%
27.0%
n.d.
64.5%
a
R ¼ –(CH2)3NHC6H4NO2; R1 ¼ –H, –OMe; R2 ¼ –H, –OEt.
Thermal and pH-stability of p-nitroaniline functionalized
xerogels
Thermal stability of the xerogels was studied using thermog-
ravimetry (TG) as well as UV/Vis absorbance spectroscopy. The
results of the thermogravimetric analyses of the various xerogels
were very similar to one another. Fig. 7 exemplifies the general
trend observed whenheating pNA-xerogels. About4% weight loss
is observed up to a temperature of 200 ꢀC due to the desorption of
residual solvents and gases as well as the liberation of ethanol and
water from hydrolysis and condensation reactions of previously
uncondensed ethoxy groups within the silica matrix. Decompo-
sition of the organic moiety commences at approximately 300 ꢀC
(see also ESI, Table S1†). UV/Vis absorption studies of heat-
treated xerogels revealed that the optical properties of the chro-
mophore change significantly once the temperature rises above
200 ꢀC as evidenced by a significant decrease of the F-value
beyond this temperature (Fig. 8 and ESI, Fig. S1†). This drop in
the F-value is also in agreement with the change in visible
appearance of the xerogels above this temperature (Fig. 8, inset).
para-Nitroaniline moieties encapsulated within the xerogels
are resistant to strong aqueous acids such as hydrochloric acid
(HCl). The yellow colour of the xerogel powders is not affected
when they are treated with concentrated 6 M HCl solution over
extended periods of time (i.e., a few months). This clearly indi-
cates that the dye is shielded from external influences by the silica
matrix despite its porosity. In contrast, the colour intensity of the
‘‘unprotected’’ pNA-chromophore obtained after dissolution of
the xerogel in aqueous NaOH is strongly pH-dependent as
shown in Fig. 9. However, the actual wavelength of the UV/Vis
absorbance maximum, lmax, is almost unaffected by the pH value
since only a small hypsochromic shift of 5 nm can be observed for
lmax when the pH changes from strongly basic to strongly acidic
conditions (see ESI, Fig. S2†). The dissolved chromophore is
accessible for protonation at the aniline nitrogen atom, which
neutralizes the colour-causing push–pull system of the
Fig. 5 13C-{1H}-CP-MAS NMR spectrum of a xerogel synthesized
according to path B of Scheme 3 [TEOS used as solvent during the SNAr-
reaction, 4-FNB : APS : TEOS ¼ (1 : 1 : 1) +9/150 ꢀC/5 h] and assign-
ment of the 13C NMR signals to the molecular structure [59 912 scans].
obtained for the as-synthesized xerogels originate from a higher
‘‘real’’ molar mass observed for the xerogels as discussed above,
i.e. due to the incomplete sol–gel process, incomplete conversion
of SNAr-reaction, or maybe a loss of the organic chromophore
during the long extraction with acetone. Table 4 further lists the
specific surface area values of the xerogels as calculated from
nitrogen adsorption measurements using the Brunauer–Emmett–
Teller (BET) method.
Lastly, the morphology of selected xerogels was studied by
electron microscopy techniques. Typical scanning electron
microscopy (SEM) and transmission electron microscopy (TEM)
images of as-obtained xerogels are shown in Fig. 6. Generally,
spherical particles approximately 1 mm in size were obtained,
although it must be noted that the particles did not exhibit
a strictly monodisperse size distribution. The spherical particles
tend to adhere to each other to form larger aggregates.
Table 4 Variation of para-nitroaniline xerogel syntheses and corresponding mass related extinction coefficients (F) as calculated from UV/Vis
absorption measurements, specific surface areas (SBET) as calculated from N2-sorption measurements and the number of experiments (n) performed
Route
EtOH : water
F/L gꢁ1 cmꢁ1
SBET/mꢁ2 gꢁ1
N
Route A (without solvent during SNAr-reaction, 4-
FNB : APS : TEOS ¼ (1 : 1 : 0) + 10/130 ꢀC/5 h)
Route B (one equivalent TEOS as solvent during SNAr-reaction,
4-FNB : APS : TEOS ¼ (1 : 1 : 1) + 9/130 ꢀC/5 h)
Route B + higher temperature (one equivalent TEOS as solvent
during SNAr-reaction, 4-FNB : APS : TEOS ¼ (1 : 1 : 1)
+ 9/150 ꢀC/5 h)
10 : 1
3 : 1
10 : 1
18.7 ꢂ 1.2
18
19.4 ꢂ 0.9
440 ꢂ 27
400
426 ꢂ 80
3
1
9
10 : 1
30 : 1
20.2 ꢂ 1.0
370 ꢂ 38
4
1
20
340
3846 | J. Mater. Chem., 2012, 22, 3839–3852
This journal is ª The Royal Society of Chemistry 2012

View Article Online
Fig. 6 Electron micrographs of a xerogel synthesized according to pathway A of Scheme 3 [no solvent during the SNAr-reaction, 4-FNB : APS :
TEOS ¼ (1 : 1) + 10/130 ꢀC/5 h]. (A–C) Scanning electron microscopy images at various magnifications and (D) transmission electron microscopy image.
particles with the chromophoric trimethoxysilane resulted in
chromophoric core–shell particles containing a solid silica core
(white) and a yellow coloured xerogel shell (Scheme 4).
Electron microscopy of the yellow powders revealed uniform
particles with diameters of 180–190 nm (Fig. 10), which is
significantly smaller than the xerogel particles obtained via our
conventional route (1 mm). The mass related extinction
Fig. 7 Thermogravimetric analysis of a xerogel synthesized according to
path B of Scheme 3 [TEOS used as solvent during the SNAr-reaction, 4-
FNB : APS : TEOS ¼ (1 : 1 : 1) + 9/150 ꢀC/5 h]. TG analysis was per-
formed in air between 40 and 900 ꢀC [heating rate: 10 K minꢁ1].
chromophore (Fig. 9, inset). This also explains why the colour of
the model compound M1 vanishes immediately when subjected
to 6 M HCl solution.
Increasing the monodispersity of p-nitroaniline functionalized
xerogels
The industrial application of pNA-functionalized xerogels as
pigments requires a high degree of monodispersity of the parti-
cles with respect to their size. A well-studied process that
Fig. 8 Thermal stability of xerogels according to their UV/Vis charac-
€
produces uniform particles in the submicron range is the Stober
terization. The mass related extinction coefficient of p-nitroaniline
xerogels, F, was determined after treating the xerogels at the indicated
temperatures for five hours under reduced pressure (5 mbar). The
observed significant decrease in F/F0 at temperatures of 200 ꢀC and above
is supported by a change in visual appearance of the sample (photographs
above each bar show the visual appearance after five hours of heat-
treatment at the indicated temperatures).
process.41 In a slight variation of our one-pot synthesis procedure
we produced the chromophoric p-nitroaniline functionalized
trimethoxysilane as described above (path B in Scheme 3), but
instead of adding the residual TEOS to the reaction mixture after
the completion of the SNAr-reaction it was used to synthesize
€
silica nanoparticles by the method of Stober. Mixing the Stober
€
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 3839–3852 | 3847

View Article Online
Fig. 9 The mass related extinction coefficient of dissolved p-nitroaniline xerogels as a function of the pH-value. The inset shows how protonation of the
aniline nitrogen atom destroys the push–pull system responsible for the yellow colour of the chromophore [R ¼ (CH2)3Si(OH)3]. F0 corresponds to the F-
value measured immediately after the dissolution of the xerogels in caustic soda.
coefficient of the core–shell xerogels was determined to be 18 L
gꢁ1 cmꢁ1, which agrees very well with that of the other 1 : 10
xerogels (Table 4) proving that the conversion is quantitative. It
is noted that the dissolution of these core–shell particles in
aqueous NaOH occurs on a much slower timescale than that of
the other xerogels, which indicates that the organically modified
In comparison, xerogels prepared by the one-pot procedure
show a hysteresis of type H2 that is typical for sol–gel derived
silica materials exhibiting irregular pore shapes.42 The specific
surface area (SBET) and pore morphologies were determined by
means of the BET and Dollimore–Heal method, respectively.
The specific surface area of the core–shell particles (32 m2 gꢁ1) is
significantly lower than that of xerogels obtained using the one-
pot procedure (567 m2 gꢁ1). The low absolute value for the core–
shell particles basically indicates that almost no porosity (i.e.,
micro/mesopores) is observed for those particles.
€
xerogel-shell shields the ‘‘Stober’’-silica core to a small extent
from external influences such as NaOH solution. Additionally,
the chromophores in the shell are still well protected against
acidic conditions as discussed previously, and no decolourization
of the particles was observed when they were stored in concen-
trated 6 M HCl solution over extended periods of time.
The expected homogeneity for the core–shell particles is also
reflected in their nitrogen adsorption–desorption isotherms, and
the type of isotherm found for these particles is typical for dense
and non-porous materials (Fig. 11).
Increasing the colour intensity of p-nitroaniline functionalized
xerogels
As elucidated in the introduction the application of para-nitro-
aniline functionalized xerogels as pigments calls for highly stable
€
Scheme 4 Synthetic concept for the preparation of chromophoric core–shell particles. (i) Sol–gel process according to Stober; (ii) SNAr-reaction in
TEOS as solvent corresponding to step 1 of both pathways in Scheme 2.
3848 | J. Mater. Chem., 2012, 22, 3839–3852
This journal is ª The Royal Society of Chemistry 2012

View Article Online
array of pNA-functionalized xerogels using APS, DAS, and TAS
as aminotrialkoxysilanes under variation of the amino-
trialkoxysilane : TEOS ratio from 1 : 10 over 1 : 5 to 1 : 1 and
analyzed the impact of both parameters on the F-value as well as
some other physico-chemical properties of the xerogels. Chart 1
shows the chemical structures of the three aminotrialkoxysilanes
used as well as those of their respective model substances, which
were synthesized to calculate the Fmax value for each xerogel
variation and simplify data interpretation. Molecular weight,
absorbance maximum, molar extinction coefficient, and F-value
are given next to the structure of each model compound.
Fmax values given next to the structures of the amino-
trialkoxysilanes correspond to values for xerogels with a amino-
trialkoxysilane : TEOS ratio of 1 : 10, 1 : 5 or 1 : 1 from top to
bottom, in that order. In the case of DAS and TAS xerogels the
values given in brackets refer to the Fmax values of mono- or
mono-/di-substitution products, respectively.
Fig. 10 (A) SEM and (B) TEM images of core–shell particles [scale bar:
1 mm and 180 nm, respectively].
and monodisperse xerogels, which are able to withstand a variety
of chemical and mechanical stresses, while at the same time
maximum colour intensity needs to be realized using minimal
quantities of the materials. Increasing the chromophore
content of xerogels can be accomplished in two ways: (i) by
reducing the TEOS amount used during the one-pot synthesis or
(ii) by replacing APS, which has only one amino group, with
trialkoxysilanes bearing two or more amino groups such as 3-(2-
aminoethylamino)propyltrimethoxysilane (DAS) and (3-trime-
thoxysilylpropyl)diethylenetriamine (TAS). We synthesized an
Although the colour intensity (hence the F-value) increases
with increasing number of available amino groups (Table 5), the
experimental F-values differ quite significantly from the theo-
retical Fmax values displayed in Chart 1. We therefore assume
that not all available amino groups of DAS and TAS are con-
verted during the SNAr-reaction and only monosubstitution is
achieved in the case of DAS and di-substitution in the case of
TAS. This hypothesis is supported by the almost similar F-values
for DAS and APS xerogels and the close to twice as high F-values
for TAS xerogels. Even longer reaction times and higher reaction
ꢀ
temperatures such as 20 h at 160 C for the reaction of 4-FNB
with DAS and 90 h at 180 ꢀC for the reaction of 4-FNB with TAS
did not significantly increase the F-values of the final products. It
seems that the substitution reaction of the secondary amine
group of DAS is sterically hindered if the primary group (from
which it is only separated by one ethylene group) has already
been converted to a para-nitroaniline moiety.
However, in the case of TAS being the nucleophile, di-
substitution at the additional secondary amine group (which is
closest to the silicon atom) is possible. Here, prior conversion of
the primary group to a pNA moiety causes less steric hindrance
because it is further away. Further, it is noted that the dissolution
of DAS- and TAS-xerogels in caustic soda solution takes much
longer than that of APS-based xerogels and in some cases
complete dissolution cannot be achieved at all. This is, however,
a critical observation since determination of the F-value via the
method described above fails in the latter case and the measured
F-value will be underestimated even if only a minor amount of
the xerogel does not dissolve. In the cases where complete
dissolution of the monosubstituted DAS-based xerogels was
achieved the slower dissolution rate by still giving similar
F-values shows that the organic chromophore of monosubtituted
DAS-based xerogels is better protected against aggressive
chemicals than that of APS-derived xerogels. Solid state
13C-{1H}-CP-MAS NMR spectroscopy cannot distinguish
between mono- and disubstitution of DAS xerogels or mono-, di-
, and trisubstitution of TAS xerogels due to the overlapping
NMR signals in combination with the lower resolution of solid
state 13C NMR spectroscopy compared to liquid NMR tech-
niques (see ESI, Fig. S3 and S4†). Elemental analysis was not
able to reveal the proper substitution state either because the
differences in nitrogen and carbon contents between mono-, di-,
Fig. 11 Nitrogen adsorption and desorption isotherms of (A) core–shell
particles and (B) of a xerogel obtained following synthesis procedure B as
described in Scheme 3 [TEOS used as solvent during the SNAr-reaction,
4-FNB : APS : TEOS ¼ (1 : 1 : 1) + 9/130 ꢀC/5 h].
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 3839–3852 | 3849

View Article Online
Chart 1 Chemical structures of mono-, di-, and tri-functional aminotrialkoxysilanes [APS: aminopropyltrimethoxysilane, DAS: 3-(2-amino-
ethylamino)propyltrimethoxysilane, TAS: (3-trimethoxysilylpropyl)diethylenetriamine] and corresponding model compounds M1–M3. Experimental
UV/Vis spectroscopic characteristics of model compounds were obtained in DMSO [3: molar extinction coefficient, lmax: wavelength of absorbance
maximum, F-value: mass specific extinction coefficient, M: molecular weight]. Maximum possible F-values (Fmax) of APS-, DAS-, and TAS-derived
xerogels were predicted based on aminosilane : TEOS ratios used (i.e., 1 : 10, 1 : 5, and 1 : 1) and F-value of the corresponding model compound [values
given in brackets refer to Fmax values for mono- or di-/mono-substituted products when using DAS or TAS, respectively].
and trisubstitution products are too small to confidently judge
the rate of substitution. The previously described incomplete
hydrolysis/condensation of TEOS during the sol–gel process
skews the results of any elemental analysis even further.
accompanies the use of DAS or TAS as aminotrialkoxysilane.
More precisely, the decomposition temperature of the organic
moiety of the xerogels merely drops from 315 ꢀC for APS-derived
ꢀ
xerogels to 297 C for TAS-derived xerogels.
The majority of the xerogels that were obtained using DAS or
TAS as aminotrialkoxysilane showed a similar morphology to
APS-derived xerogels in that they consisted of spherical particles
approximately 1 mm in diameter. The specific surface area of
DAS- or TAS-based xerogels as measured by N2-adsorption is
very similar to that of the APS-derived powders for all tri-
alkoxysilane : TEOS ratios studied (Table 5). Other physico-
chemical properties of the xerogels such as their thermal stability
are nearly unaffected by the increase in organic content that
The colour intensity obviously also increases with decreasing
TEOS content of the reaction mixture (Table 5), but again, the
theoretical Fmax values were not achieved (Chart 1). A decrease in
the amount of TEOS used during the synthesis produces smaller
specific surface areas due to the lowered porosity of the materials.
As shown in Table 5, variation of the amino-
trialkoxysilane : TEOS ratio has a much larger effect on the
specific surface area than variation of the number of available
amino groups.
Table 5 Mass related extinction coefficients (F) and specific surface areas (SBET) of xerogels derived from mono-, di-, and tri-functional amino-
trialkoxy-silanes [APS, DAS, TAS], 4-FNB, and TEOS [averaged values of multiple batches (n) are given]
APS
DAS
1 : 10
TAS
Silanea
Silane : TEOSb
1 : 10
1 : 5
1 : 1
1 : 5
1 : 1
1 : 10
1 : 5
1 : 1
c
c
F/L gꢁ1 cmꢁ1
SBET/m2 gꢁ1
n
19 ꢂ 1
410 ꢂ 65
18
28 ꢂ 1
120 ꢂ 50
3
61 ꢂ 1
21 ꢂ 2
250 ꢂ 50
6
32 ꢂ 2
110 ꢂ 50
7
(>55)c
<2
5
(>39)c
260 ꢂ 30
2
—
—
<2
2
80
1
<2
4
a
b
trialkoxysilane; molar ratio of trialkoxysilane and TEOS; c not completely soluble in 1 M caustic soda solution.
3850 | J. Mater. Chem., 2012, 22, 3839–3852
This journal is ª The Royal Society of Chemistry 2012

View Article Online
4 S. C. Klein, Sol–Gel Optics, Processing and Applications, Kluwer
Academic Press, Boston, 1994.
5 C. Sanchez and F. Ribot, New J. Chem., 1994, 18(10), 1007–
1047.
The yellow-coloured pNA-xerogels were previously tested as
pigments for the colouration of dental composite materials based
on methacrylates which are commonly used for the restoration of
teeth.43 The mass of pigment required to obtain a desired col-
ouration using the organic–inorganic xerogels is reduced by 96%
compared to the formerly used pigment Yellow 53 (Lichtgelb
8G) resulting in improved optical properties of the composite
such as transparency as well as superior mechanical properties.44
6 Z. Yang, C. Xu, B. Wu, L. R. Dalton, S. Kalluri, W. H. Steier, Y. Shi
and J. H. Bechtel, Chem. Mater., 1994, 6(10), 1899–1901.
7 D. Avnir, Acc. Chem. Res., 1995, 28(8), 328–334.
8 U. Schubert, N. Huesing and A. Lorenz, Chem. Mater., 1995, 7(11),
2010–2027.
9 K. Yagi, S. Shibata, T. Yano, A. Yasumori, M. Yamane and
B. Dunn, J. Sol-Gel Sci. Technol., 1995, 4(1), 67–73.
10 F. Chaput, D. Riehl, J. P. Boilot, K. Cargnelli, M. Canva, Y. Levy
and A. Brun, Chem. Mater., 1996, 8(2), 312–314.
11 C. Sanchez and B. Lebeau, Pure Appl. Opt., 1996, 5(5), 689–699.
12 J. Biteau, F. Chaput, K. Lahlil, J.-P. Boilot, G. M. Tsivgoulis,
J.-M. Lehn, B. Darracq, C. Marois and Y. Levy, Chem. Mater.,
1998, 10(7), 1945–1950.
Conclusions
Highly thermally and chemically stable, yellow-coloured xerogels
were synthesized by combination of nucleophilic aromatic
substitution reaction of 4-FNB with APS and a subsequently
induced sol–gel process with TEOS as co-condensing silica
source in a one-pot synthesis. Reaction parameters such as
temperature, reaction time, concentrations and sol–gel condi-
tions are systematically studied to optimize the conversion rate
and reproducibility. We also prove for the first time that the HF
formed during the SNAr-reaction readily reacts with the silanes
present in the reaction mixture, since it cannot be trapped by
acid–base reactions, and catalyzes the subsequent sol–gel
process. The materials can withstand temperatures of up to 200
ꢀC and strongly acidic pH-conditions without any loss of colour.
The chromophore is embedded in the silica matrix and therefore
not accessible to strong acids that immediately decolourize
a corresponding model compound by protonation. Further,
core–shell particles with solid silica cores and chromophoric
shells were produced. They exhibit the same dye content as our
conventional xerogels, but combine it with more desirable
monodisperse morphology and smaller particle diameters in the
range of 200 nm. Finally, two convenient methods to enhance the
colour intensity of the materials are presented. The main
approach of increasing the number of available amino groups in
the n-aminoalkyltrialkoxysilane leads to an increase in colour
intensity (although steric hindrance prohibits full conversion of
all amino groups) and enhanced chemical robustness, while other
physicochemical properties remain almost identical. On the other
hand, the far more obvious approach of decreasing the TEOS
amount used in the synthesis also increases the colour intensity,
but is accompanied by a drastic alteration of other properties
such as the specific surface area.
13 H. Jiang and A. K. Kakkar, J. Am. Chem. Soc., 1999, 121(15), 3657–
3665.
14 R. J. P. Corriu, Angew. Chem., 2000, 112(8), 1432–1455.
ꢀ
15 J. A. Reyes-Esqueda, A. Franco, M. Bizarro, J. Garcıa-Macedo,
M. Canva, B. Darracq, Y. Levy, K. Lahlil, F. Chaput and
J. P. Boilot, J. Sol-Gel Sci. Technol., 2003, 26(1), 1011–1016.
16 P. Gomez-Romero and C. Sanchez, Functional Hybrid Materials,
Wiley-VCH Verlag GmbH, Weinheim, 2004.
17 J. Bujdak and N. Iyi, J. Phys. Chem. B, 2006, 110(5), 2180–2186.
€
18 A. Seifert, K. Ladewig, P. Schonherr, K. Hofmann, R. Lungwitz,
I. Roth, A. Pohlers, W. Hoyer, G. Baumann, S. Schulze,
M. Hietschold, N. Moszner, P. Burtscher and S. Spange, J. Sol-Gel
Sci. Technol., 2010, 53(2), 323–341.
€
€
19 S. Spange, A. Seifert, H. Muller, S. Hesse and C. Jager, Angew.
Chem., Int. Ed., 2002, 41(10), 1729–1732.
€
€
20 A. Seifert, S. Spange, H. Muller, S. Hesse and C. Jager, J. Sol-Gel Sci.
Technol., 2003, 26(1), 77–81.
21 K. Schreiter, K. Hofmann, A. Seifert, A. Oehlke, K. Ladewig,
€
T. Ruffer, H. Lang and S. Spange, Chem. Mater., 2010, 22(9),
2720–2729.
22 I. Roth, F. Simon, C. Bellmann, A. Seifert and S. Spange, Chem.
Mater., 2006, 18(20), 4730–4739.
23 C. R. Morcombe and K. W. Zilm, J. Magn. Reson., 2003, 162(2), 479–
486.
24 R. Joseph, S. Zhang and W. T. Ford, Macromolecules, 1996, 29(4),
1305–1312.
ꢀ
25 D. Massiot, F. Fayon, M. Capron, I. King, S. L. Calve, B. Alonso,
J.-O. Durand, B. Bujoli, Z. Gan and G. Hoatson, Magn. Reson.
Chem., 2002, 40(1), 70–76.
26 C. E. S. Alvaro and N. S. Nudelman, J. Phys. Org. Chem., 2005, 18(8),
880–885.
27 H. Grube and H. Suhr, Chem. Ber., 1969, 102(5), 1570–1579.
28 J. Sauer and R. Huisgen, Angew. Chem., 1960, 72(9), 294–315.
29 C. Claude, B. Garetz, Y. Okamoto and S. Tripathy, Mater. Lett.,
1992, 14(5–6), 336–342.
30 J. M. Allen, S. L. Aprahamian, E. A. Sans and H. Shechter, J. Org.
Chem., 2002, 67(11), 3561–3574.
31 R. Tacke, J. Becht, A. Lopez-Mras, W. S. Sheldrick and A. Sebald,
Inorg. Chem., 1993, 32(12), 2761–2766.
32 R. Tacke, J. Becht, A. Lopez-Mras and J. Sperlich, J. Organomet.
Chem., 1993, 446(1–2), 1–8.
Acknowledgements
Financial support by the Deutsche Forschungsgemeinschaft
(DFG) and the Fonds der Chemischen Industrie is gratefully
acknowledged. The authors thank M. Berger, S. Ebert, and D.
Boettger-Hiller for nitrogen adsorption measurements, G. Bau-
33 R. Tacke, A. Lopez-Mras, W. S. Sheldrick and A. Sebald, Z. Anorg.
Allg. Chem., 1993, 619(2), 347–358.
34 R. Tacke, J. Becht, O. Dannappel, R. Ahlrichs, U. Schneider,
W. S. Sheldrick, J. Hahn and F. Kiesgen, Organometallics, 1996,
15(8), 2060–2077.
€
mann and S. Schulze for TEM/SEM imaging, and P. Schonherr
35 A. Adima, C. Bied, J. Moreau and M. Wong Chi Man, Eur. J. Org.
Chem., 2004, (12), 2582–2588.
36 R. J. P. Corriu, C. Guerin, B. J. L. Henner and W. W. C. Wong Chi
Man, Organometallics, 1988, 7(1), 237–238.
for assistance with DSC/TGA measurements.
37 J. A. Deiters and R. R. Holmes, J. Am. Chem. Soc., 1990, 112(20),
7197–7202.
Notes and references
38 R. R. Holmes, Chem. Rev., 1990, 90(1), 17–31.
39 C. J. Brinker and G. W. Scherer, Sol–Gel Science: the Physics and
Chemistry of Sol–Gel Processing, Academic Press, San Diego, 6th
edn, 1995.
1 R. J. Jeng, Y. M. Chen, A. K. Jain, J. Kumar and S. K. Tripathy,
Chem. Mater., 1992, 4(5), 972–975.
2 J. Kim, J. L. Plawsky, R. LaPeruta and G. M. Korenowski, Chem.
Mater., 1992, 4(2), 249–252.
40 G. Hartmeyer, C. Marichal, B. Lebeau, P. Caullet and J. Hernandez,
J. Phys. Chem. C, 2007, 111(18), 6634–6644.
3 F. Chaput, D. Riehl, Y. Levy and J. P. Boilot, Chem. Mater., 1993,
5(5), 589–591.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 3839–3852 | 3851

View Article Online
€
41 W. Stober, A. Fink and E. Bohn, J. Colloid Interface Sci., 1968, 26(1),
62–69.
43 N. Moszner and S. Klapdohr, Int. J. Nanotechnol., 2004, 1(1–2), 130–
156.
42 S. J. Gregg and K. S. Sing, Adsorption, Surface Area and Porosity,
Academic Press, London, 2nd edn, 1995.
44 S. Spange, N. Moszner, P. Burtscher and V. Rheinberger, Eur. Pat.
Appl., DE 101 373 72, EP 1 281 383 B1, US 6,833,394 B2, 2003.
3852 | J. Mater. Chem., 2012, 22, 3839–3852
This journal is ª The Royal Society of Chemistry 2012