Materials Chemistry and Physics 140 (2013) 168e175

Contents lists available at SciVerse ScienceDirect

Materials Chemistry and Physics

journal homepage: www.elsevier.com/locate/matchemphys

Novel silica/bis(3-aminopropyl) polyethylene glycol inorganic/organic hybrids

by solegel chemistry
Laura Russo a, Luca Gabrielli a, Esther M. Valliant b, Francesco Nicotra a, Jesús Jiménez-Barbero c,
Laura Cipolla a, *, Julian R. Jones b, **
a
Dept. of Biotechnology and Biosciences, University of Milano-Bicocca, P.za della Scienza 2, 20126 Milano, Italy
b
Department of Materials, Imperial College London, South Kensington Campus, London SW7 2AZ, UK
c
Centro de Investigaciones Biológicas, CSIC, Ramiro de Maeztu 9, 28040 Madrid, Spain

h i g h l i g h t s

New solegel hybrids were developed.

Novel silica/bis(3-aminopropyl) polyethylene glycol inorganic/organic hybrids.

NMR was used to characterise the fabrication process.

The novel materials showed interesting mechanical properties.

a r t i c l e i n f o a b s t r a c t

Article history: Biomaterials are needed for tissue regeneration applications that provide control of mechanical prop-
Received 4 December 2012 erties and enhanced toughness compared to conventional bioceramics. New solegel hybrids were
Received in revised form developed with interpenetrating networks of silica and bis(3-aminopropyl) polyethylene glycol. Covalent
28 February 2013
coupling between the organic and the inorganic components was used to control mechanical properties
Accepted 8 March 2013
of the hybrids. The objective was to synthesise and characterise a bis(3-aminopropyl) polyethylene glycol
silica hybrid material with 35 wt% organic and 65 wt% inorganic and covalent coupling between the
Keywords:
components. A coupling agent, 3-glycidopropyltrimethoxysilane (GPTMS) was used to form the covalent
Biomaterials
Composite materials
links. The hypothesis was that the epoxy ring of the GPTMS would react with the polymer, leaving a
Chemical Synthesis polymer functionalised with siloxane groups. In a sol of hydrolysed tetraethyl orthosilicate (TEOS) the
NMR siloxanes from the GPTMS form eSieOeSie bonds between the functionalised polymer and the silica
network. Bis(3-aminopropyl) polyethylene glycol contains two terminal amino groups available for the
covalent functionalisation with the epoxy group of GPTMS. Hybrids with 35 wt% organic and 65 wt%
inorganic with a ratio of GPTMS:PEG of 1:4 were proven to have an excellent balance between strain to
failure (10%) and compressive strength (20 MPa). However, the functionalisation of the polymer was
followed by liquid NMR as a function of the aging time and temperature and the expected reaction of
nucleophilic attack of the epoxy ring by the amino group of the polymer did not happen until the water
was removed from the system during drying. Increasing the amount of GPTMS decreased rate of weight
loss during immersion in TRIS buffer solution.
Ó 2013 Elsevier B.V. All rights reserved.

1. Introduction However, they are brittle and to overcome this drawback, inor-
ganiceorganic hybrids have been synthesised to improve tough-
Bioactive glass can bond with bone and stimulate new bone ness. The homogeneous combination of inorganic and organic
growth, making them good candidates for tissue engineering [1,2]. moieties in a single-phase material allows the tuning material
properties in respect to the desired application [3e7]. Our work on
* Corresponding author. Tel.: þ39 0264483460; fax: þ39 0264483565. solegel glass scaffolds has shown that bioactive glasses can be made
** Corresponding author. Tel.: þ44 20 7594 6749. to have ideal structures for bone regeneration [8]. Hybrid bio-
E-mail addresses: laura.russo@unimib.it (L. Russo), l.gabrielli2@ materials may have interesting properties: for example organice
campus.unimib.it (L. Gabrielli), e.valliant08@imperial.ac.uk (E.M. Valliant),
francesco.nicotra@unimib.it (F. Nicotra), jjbarbero@cib.csic.es (J. Jiménez-Barbero),
inorganic hybrids may be made into 3D macroporous scaffolds with
laura.cipolla@unimib.it (L. Cipolla), julian.r.jones@imperial.ac.uk (J.R. Jones). appropriate 3D pore networks for tissue regeneration [8,9]. Hybrids

0254-0584/$ e see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.matchemphys.2013.03.016
 L. Russo et al. / Materials Chemistry and Physics 140 (2013) 168e175 169

Fig. 1. Organic and inorganic reagents for hybrid synthesis: GPTMS ¼ 3-glycidoxypropyltrimethoxysilane; dPEG ¼ bis(3-aminopropyl) polyethylene glycol; TEOS ¼ Tetraethyl
orthosilicate.

are usually synthesised by the solegel process, where the organic and silica networks. The primary aim was to produce hybrids with
part is introduced early in the process so that an inorganic (silica) improved mechanical properties to bioactive glasses and bio-
network forms around the polymer molecules [10]. The solegel ceramics. Tetraethyl orthosilicate (TEOS), GPTMS and dPEG (Fig. 1)
technique has been used to produce several types of hybrid mate- were used as components of the reaction system. The bis(3-
rials for biomedical [11e18] and non biomedical applications aminopropyl) polyethylene glycol contains two amino terminal
[3,19,20]. The synthesis is however relatively complex and there are groups available for the reaction with the epoxy group of GPTMS.
several chemistry challenges that must be overcome before hybrids The hypothesis was that a dPEG/silica hybrid would have
will be successful for tissue engineering application, e.g. incorpo- enhanced mechanical properties compared to a bioactive glass and
ration of a suitable polymer and fulfilling the potential of controlled that during the functionalisation of the dPEG, the epoxy ring of the
degradation and mechanical properties. GPTMS would open and a covalent bond would form between the e
Polyethylene glycol (PEG) is a promising polymer for solegel NH2 groups on dPEG and GPTMS (nucleophilic attack). The siloxane
hybrid synthesis as it is biocompatible and biodegradable. PEG/silica pendant groups on dPEG hydrolyse upon immersion in the sol and
hybrids were recently produced by Liu et al. [21] where they added the newly formed SieOH groups then condense with the silica
the PEG into the TEOS-based sol at pH 2. No covalent bonds were network that forms from the condensation of hydrolysed TEOS. The
formed between the components. The rate of loss of PEG from the aim was to investigate this hypothesis.
hybrid was not assessed, but is likely to be rapid due to the solubility
of PEG and the weak interactions between the components.
One way to obtain stronger interaction between the inorganic and 2. Materials and methods
organic components in the hybrid material and gain tight control over
the degradation rate and mechanical properties to obtain covalent 2.1. Hybrids synthesis
coupling between the organic and inorganic components e.g. through
silane coupling agents such as 3-glycidoxypropyltrimethoxysilane All reagents were supplied by Sigma Aldrich, UK unless other-
(GPTMS). GPTMS is one of the most common precursors for the wise specified. Fig. 2 shows a flow chart of the synthesis. Initially
preparation of hybrid organiceinorganic materials [22]. The chem- 0.500 g of dPEG (Mw 1500 Da) was dissolved in 2 mL of deionised
istry of GPTMS is quite complex and aspects are not fully explored and H2O. This solution was functionalised by addition of an appropriate
understood [23]. One of the main difficulties in controlling the amount of GPTMS (the molar ratio of dPEG/GPTMS) of 1:2 (dPEG:
chemical reactions during the solegel process of GPTMS is that 0.500 g; GPTMS: 0.144 mL) and 1:4 (dPEG: 0.500 g; GPTMS:
several competitive reactions can occur. Usually, when in solution, the 0.289 mL). The functionalised dPEG solution was left to mix (2 h)
epoxy ring of GPTMS (Fig. 1) can be opened by water (affording the before being added as a solegel precursor. A sol was prepared by
corresponding diol), or by nucleophilic attack by eCOOH or eNH2 mixing 592.2 mL of deionised H2O with 0.197 mL of HCl (1 N) fol-
containing molecules [10], and a bond between the GPTMS and the lowed by 1.903 mL of tetraethyl orthosilicate (TEOS). The molar
organic molecule is formed. In the presence of amines as the nucle-
ophiles, GPTMS affords a functionalised trimethoxyorganosilane that
can be polymerised with the solegel derived inorganic component,
e.g. silica species arising from the hydrolysis of tetraethyl orthosilicate
(TEOS), through the condensation of SieOH groups to form eSieOe
Se bonds [24]. The simultaneous polymerisation of the inorganic and
organic network is a competitive process [25,26]; a faster poly-
condensation rate of the inorganic network hinders the polymerisa-
tion of the organic network and vice versa [17]. The strict control of
the competitive reactions is never simple since small changes in the
synthesis protocol, e.g. pH, can affect materials properties such as
connectivity of the silica network, degree of covalent coupling and
mechanical properties [27,28]. To be able to tune the physical and
mechanical property of the hybrid material it is important to un-
derstand the reactivity of GPTMS with the organic component.
Polyfunctional amines are one of the most popular curing agents
that are commonly used to fabricate a wide class of thermosetting
organic polymers based on the epoxy-amine system. The reactions
between epoxy and amino groups to obtain homogeneous mate-
rials have been the object of extensive research in industry [29]. In
the present work we propose the fabrication of novel polyoxy-
ethylene bis(amine)/silica (dPEG/silica) solegel hybrids of 35 wt% Fig. 2. Flow chart showing the reaction and processing steps involved in the synthesis
organic and 65 wt% silica with covalent coupling between the dPEG of the dPEG/silica hybrid.
170 L. Russo et al. / Materials Chemistry and Physics 140 (2013) 168e175

ratio of water/TEOS (R ratio) was 4 and HCl 2 N was added to
catalyse TEOS hydrolysis at a volume ratio (water/HCl) of 3. The
solution was stirred (1 h) to allow hydrolysis of TEOS, before adding
the functionalised dPEG solution, which was mixed for a further
1 h. Once the functionalised dPEG solution has mixed for 1 h in the
sol, aliquots of the sol (4 mL) were then transferred into screw top
teflon moulds (3 cm diameter) for gelling and aging. The gels were
aged (3 days at 40  C) and then dried (8 days at 60  C).
The 5 mm diameter NMR tube was sealed and maintained closed
during the full duration of the experiment. The number of scans was
between 160 and 240, depending on the signal-to-noise ratio, and
the recycling delay was set to 5 s. 13C NMR spectra were recorded by
using a 200 ppm spectral width, using between 200 and 1268 scans
depending on the signal-to-noise ratio.
2.6. Degradation rate studies

2.2. Hybrid characterisation TRIS-buffer solution was made following the procedure
described by Brauer et al. [30]. PEG:GPTMS 1:2 and 1:4 (75 mg) in
ATR/FT-IR absorption spectra (Perkin Elmer Spectrum 100) were triplicates were immersed in 50 mL of TRIS-buffer at 37  C and
recorded on five samples for each type of functionalisation in the
range 4000e250 cm1, scan time 20, at a resolution of 4 cm1. Field
emission gun scanning electron microscopy (LEO Gemini 1525
FEGSEM) of hybrid nanostructure was carried out on gold coated
hybrids to image their nanoscale morphology, with an accelerating
voltage of 5 kV and a working distance w7 mm.

2.3. Compression testing

Compression testing was carried using an Instron 5866 on

rectangular samples with height:width ratio of 2:1 (n ¼ 5). A 500 N
load cell was used for testing, with a compression extension speed
of 0.5 mm min1. At five samples (n ¼ 5) of each composite system
were measured. The mean and standard deviation were reported.
The modulus (taken at the maximum modulus before the onset of
yield), yield stress (defined as the end of the linear deformation
region) and the corresponding strain at the yield point were also
recorded. The yield strength was determined from the cross point of
the two tangents on the stressestrain curve around the yield point.

2.4. Liquid NMR experiments

Benzyl alcohol was used as internal standard for quantification

of the GPTMSedPEG reaction. The pulse delay was sufficient to
prevent saturation. Chemical shifts were referenced externally to
TMS at 0 ppm.
This molecule has a high boiling point, it does not have functional
groups that could influence the possible reactions, neither at the
epoxide moiety nor at the silica network formation, and its 1H NMR
signals do not overlap to those of the reagents. Using the integral of the
aromatic protons of benzyl alcohol (HePh, m, 5H, d 7.48e7.31 ppm),
and referring that to the CH2 group closest to the silicon atom (H-1, m,
2H, d 0.87e0.60 ppm) and those assigned to the NH2CH2CH2CH2
moiety (H-b, m, 4H, d 2.01e1.89 ppm) of dPEG, the different compo-
nents were normalised, and the amount of GPTMS and dPEG present
at any time in solution was thus estimated.

2.5. Hybrid synthesis

450 mg of dPEG (0.3 mmol) was dissolved under stirring in

2.55 mL of D2O. The pH of the solution was adjusted to 6 by adding a
DCl 2 N solution in D2O, then 16 ml of benzyl alcohol (0.154 mmol)
were added for the experiments requiring internal standard. 130 ml
(0.6 mmol) or 260 ml (1.2 mmol) of GPTMS (respectively for
PEG:GPTMS 1:2 and 1:4 experiments) were added and the solution
was stirred for 2 h at room temperature, in a sealed vial. After this
stage, the vial was transferred to an oven at 40  C, for 72 h, without
stirring. After 2 h at room temperature and after 24, 48 and 72 h at
40  C, 400 ml of the reaction were collected, diluted adding 100 ml of
D2O, and the resulting solution was analysed by NMR spectroscopy,
13 1
C, H and measurements were carried out using a Varian 400 MHz
Mercury instrument, operating at a proton frequency equal to Fig. 3. 1H NMR spectra and signal assignments of a) GPTMS in D2O, b) GPTMS D2O/DCl
400 MHz; chemical shifts were referenced to methanol (d 3.40 ppm). at pH 2, c) dPEG in D2O.
 L. Russo et al. / Materials Chemistry and Physics 140 (2013) 168e175 171

stirred at 120 rpm. At 1, 4, 24, 72, 168, 336, 504 and 672 h aliquots of
1.5 mL of the solution were removed and replaced with 1.5 mL of
fresh TRIS-buffer solution.
3. Results and discussion
Two different ratios of GPTMS and dPEG were combined with a
silica precursor to obtain Class II hybrid materials. The interaction
between the organic and inorganic components, and the formation
of covalent coupling reaction was examined, specifically the nature
of the covalent coupling reaction and the formation of the silica
network. For coupling to be successful, GPTMS must react both
with the polymer and the silica network. The timing of the
condensation of the siloxane groups and the reaction of the organic
moiety with the GPTMS is a crucial issue for the formation of a
successful hybrid, influencing its degradation rate and mechanical
properties. In order to better understand the reaction system dur-
ing the functionalisation and the aging process liquid Nuclear
Magnetic Resonance (NMR) was efficiently used. Aging is the
consolidation of a gel in a sealed container post-gelation, prior to
the drying step [31]. 29Si and 13C solid state NMR have been widely
used in the literature to characterise solid solegel materials [32e
34]. Solution 1H and 13C NMR was used to study the reaction be-
tween GPTMS and bis(3-aminopropyl) polyethylene glycol (dPEG)
in order to characterise the competition of the organic versus
inorganic polymerisation.

Fig. 4. A: Solution phase 1H NMR of the reaction between dPEG/GPTMS, ratio 1:2; B: solution phase 1H NMR of the reaction between dPEG/GPTMS, ratio 1:4.
172 L. Russo et al. / Materials Chemistry and Physics 140 (2013) 168e175

Fig. 5. Amount variations of dPEG and GPTMS in the two reaction systems. For clarity, arabic numbers are used to designate protons in the GPTMS, while greek letters are used for
dPEG protons.

The two components (dPEG and GPTMS) were reacted,
following a solegel protocol, as outlined schematically in Fig. 2. In
order to obtain hybrids with different properties, two different
ratios between the GPTMS and the dPEG were used (dPEG:GPTMS
1:2 and 1:4). The sol was prepared by the hydrolysis of the silica
precursor tetraethyl orthosilicate (TEOS) in acidic conditions (pH
w1). In a separate beaker, a mixture of dPEGeGPTMS was prepared
at pH ¼ 6. Afterwards, the TEOS solution was added into the dPEGe
GPTMS mixture [10]. The complete sol was poured into Teflon
moulds and sealed for the aging at 40  C for 3 days (Step 4, Fig. 2),
after which the tops were unscrewed for a final drying step (Step 5,
Fig. 2) at 60  C for 8 days.
In order to characterise the polymerisation reaction between
the organic moieties of the system, that is GPTMS epoxy group and
amino groups of the dPEG, a solution-state NMR study was per-
formed; as far as we know, this is the first time that liquid NMR
experiments have been performed on such a system. Homo- and
heteronuclear 2D NMR experiments of GPTMS (Fig. 3a), and GPTMS
in D2O/DCl at pH 2 (Fig. 3b) and dPEG in D2O (Fig. 3c), were
recorded and used as reference spectra.
The GPTMSedPEG reaction at pH 6 was monitored by NMR at
different reaction times. 1H NMR spectra of the reactions in
deuterated water were recorded at the very end of the 2 h func-
tionalisation step (Step 1 in Fig. 2) at room temperature and during
the aging step (Step 4 in Fig. 2) at different times (24, 48, and 72 h at
40  C).
Interestingly, for both the 1:2 and 1:4 dPEG:GPTMS mixtures,
after 2 h at room temperature, no reaction had occurred at the
epoxide ring (H6a d 2.75, H6b d 2.94, Fig. 3a). In the functionalisation
step (Step 1 in Fig. 2), the only change detected was partial hydrolysis
of the silicon methoxy groups, resulting in the formation of meth-
anol (signal of the MeOH at d 3.34 ppm) and some SieOH groups.
Thus, the reaction between the epoxide and the dPEG did not occur
at this early stage. After 24 h of the aging step at 40  C (Step 4 in
Fig. 2), the NMR data (following the normalisation procedure
described at the experimental section) showed that the methoxy
groups of the GPTMS were fully hydrolysed (Fig. 4). In particular, the
signals due to H2 (1.84e1.60 ppm) and H1 (0.87e0.60 ppm) of
GPTMS were much broader and asymmetric, compared to those
recorded after 2 h reaction at room temperature. By 24 h, for both

Fig. 6. FT-IR spectra of: dPEG, dPEG/silica hybrids with ratios of dPEG:GTPMS of 1:2 and 1:4, and GPTMS/TEOS samples after drying, A) wavenumbers 250e4000 cm1 and B)
expanded view of 1300e1750 cm1.
 L. Russo et al. / Materials Chemistry and Physics 140 (2013) 168e175 173

Fig. 7. Stressestrain curves obtained by compression test of A) 1:2 dPEG/GPTMS and B) 1:4 dPEG/GPTMS hybrids.

reactions (1:2 and 1:4 dPEG/GPTMS molar ratios), the formation of
solid material started to be detectable and most of the GPTMS had
reacted, forming a silica network. In contrast, almost all the dPEG
was still unreacted, as quantified by NMR (Fig. 4A and B, experi-
mental section for details). After 72 h at 40  C, the GPTMS signals had
almost disappeared, indicating that most of the GPTMS had
condensed, and the hybrid had gelled.
As previously mentioned, after 24 h of the aging step, formation
of solids occurred, due to reaction of the species in solution. In
order to better characterise the reaction system, using the internal
standard (benzyl alcohol) we decided to quantify the reacting
species of the system. In fact, the integrals of the peaks of protons of
dPEG (greek letters) and of GPTMS derivatives (arabic numbers) can
be used to unambiguously quantify the extent of the cross-reaction
involving the two species in solution as a function of time. In
particular, H-b (amine functional group) of dPEG (d 2.01e1.89 ppm)
was chosen as characteristic PEG’s peak because it doesn’t overlap
to other signals; H-1 (alkoxysilane group) of GPTMS (d 0.87e
0.67 ppm) was chosen because is the only non overlapping signal
common to all the possible GPTMS derivatives that can be formed
during the reaction (condensation products, epoxide hydrolysis to
the corresponding diol, cross-reaction of the epoxide with the
amine nucleophile). A decrease of the integrals indicates that the
concentration of the compound in solution is decreasing due to
solids formation; on the contrary a constant value indicates that the
compound does not give cross-reaction with the other reagent,
being constantly present in solution, despite the observed solid
formation. The variation of the integrals values as a function of
reaction time is plotted in Fig. 5: the integral value for the dPEG
proton H-b in solution is approximately constant in both reactions
with different dPEG/GPTMS ratios (H-b integrals), indicating that
the dPEG did not react with the epoxide of GPTMS, at least in the
first 72 h. On the contrary, integrals of H-1, decreased as a function
of time; moreover no reaction on the epoxide has been detected
while the GPTMS is in solution, suggesting that GPTMS is aggre-
gating in its epoxy form and that reaction on the epoxy ring hap-
pens afterwards, as confirmed by FT-IR (vide infra).
In summary, solution-phase NMR has been extremely useful to
characterise the reaction features in first steps of the material
fabrication; furthermore, it appears clear that the amine nucleo-
phile is not reacting with the epoxy ring in solution phase. The final
hybrids were produced by drying the aged gels (Step 5 in Fig. 2) at
60  C for 8 days. The dried transparent crack-free materials were
obtained and characterised in terms of morphology and mechanical
properties and by FT-IR. FT-IR spectra (Fig. 6) of dPEG/GPTMS hy-
brids after drying with two different ratios of the organic compo-
nents (dPEG:GPTMS ratios of 1:2 and 1:4) were collected and
compared to a spectrum from a GPTMS/TEOS hybrid (without

Fig. 8. SEM micrographs of silica/PEG hybrids with dPEG:TEOS and molar ratios of dPEG/GPTMS of A) 1:2, B) 1:4.
174 L. Russo et al. / Materials Chemistry and Physics 140 (2013) 168e175
additional organic components) as a reference. The vibrational
modes due to SieOeSi asymmetric stretching were observed at
approximately 1070 and 1200 cm1. Additional peaks were present
at 790 cm1 (symmetric SieOeSi stretching vibration) and at
450 cm1 (SieOeSi bending modes). Vibration bands from the
main silica network content in the hybrid samples between 400
and 1250 cm1 were also present. The additional absorption bands
observed around 1070 cm1 were ascribed to the organic moieties
of the material, in particular, to the CeO and CeC bonds of the alkyl
silane. Other signals between 1350 and 1500 cm1 were due to the
CeN stretching vibrations present in plain dPEG and in the hybrid
samples, were absent in the reference material (TEOS/GPTMS). The
primary amino groups of the dPEG gave a characteristic broad
signal at around 1600 cm1, which was not present in the spectra of
the hybrids, confirming that these groups were involved in covalent
bonding with GPTMS [35,36].
The FT-IR data acquired on the solid material, together with the
NMR observations in solution, suggest that during the aging time at
40  C the hydrolysis of methoxy silane groups occurs, with conse-
quent silica network formation. During this step, the epoxide ring
did not undergo nucleophilic attack by the amino groups of dPEG.
However, the unambiguous disappearance of the signal of the
primary amino groups of the dPEG around 1600 cm1 in the hy-
brids indicated that the dPEG amino groups were reacted, but most
likely the reaction occurred during the drying step at 60  C.
The delay in reaction between GPTMS and dPEG until the drying
stage may be because the eNH2 group was unavailable for nucle-
ophilic attack, as it was involved in the catalysis of silica network
formation during gelation and aging [37]. This suggests that the
silica network formed first. As the drying progressed, and the acid
became more concentrated, the conditions became favourable for
nucleophilic attack of the epoxy ring.
Compressive mechanical testing was performed on the hybrid
monoliths to test the strength of the hybrids. An initial assess-
ment of the mechanical behaviour of the hybrid materials was
carried out at room temperature with a compression test with
samples prepared using the two hybrids materials with the
compositions 1:2 and 1:4 dPEG/GPTMS and that of solegel glass
70S30C monoliths (sintered at 800  C) (Fig. 7). The strain to
failure increased from 4.2% for the solegel glass to 8e15% for the
hybrid with a1:4 ratio of dPEG:GPTMS, indicating improved strain
to failure and therefore increased toughness. The compressive
strength of the hybrid monoliths (15e20 MPa) was low compared
to 70S30C (32e89 MPa) and to that of cortical bone (100e
230 MPa) [31]. The hybrid with a 1:2 ratio of dPEG:GPTMS, had
the lowest compressive strength. Increasing the covalent coupling
from 1:2 to 1:4 dPEG/GPTMS increased compressive strength and
elongation to failure.
The transparent nature of the monoliths also indicates homo-
geneity (Fig. 8). The fracture surfaces of 1:2 and 1:4 dPEG/GPTMS
hybrid materials were compared using SEM (Fig. 8). The 1:4 hybrids
exhibited a fracture surface indicative of brittle fracture (Fig. 8B),
similar to that of a glass. However the 1:2 hybrids have a more
ductile mode of failure (Fig. 8A).
A preliminary evaluation of the degradation behaviour of the
hybrids was conducted through mass loss upon immersion in TRIS
medium. Results are summarised in Fig. 9. The mass loss values for
both samples was more rapid at early time points, after which the
rate of weight loss reduced. The mass loss was higher for hybrids
with a 1:2 ratio of dPEG:GPTMS (76% at 3 weeks) compared to
hybrids with a 1:4 ratio of dPEG:GPTMS (52% at 3 weeks). Therefore
mass loss was significantly reduced when the amount of GPTMS
was increased. These results support the data that suggested co-
valent coupling did occur during drying and imply that covalent
coupling can be used to tailor stability in solution.
Fig. 9. Degradation assay into TRIS-buffer solution as a function of soaking time for
dPEG:TEOS and molar ratios of dPEG/GPTMS of A) 1:2, B) 1:4.
4. Conclusions
A new solegel hybrid material was synthesised using dPEG as a
novel cross-linking molecule. The work was initially based on the
hypothesis that reacting dPEG with GPTMS at pH 6 would yield a
polymer functionalised with GPTMS molecules, through nucleo-
philic attack of the epoxy ring by eNH2 groups, and this would
enable covalent coupling of the dPEG to the silica network. In order
to prove this working hypothesis, for the first time we introduced
solution-phase NMR as relevant tool in order to follow the reaction
during material processing. By NMR, we observed, instead, that
when the dPEG and GPTMS were mixed together, the siloxane
groups on the GPTMS hydrolysed and the resulting SieOH groups
condensed to form eSieOeSie bonds. This means the GTPMS is
likely to react with itself and the solegel silica network that forms
in the TEOS based sol when the dPEG/GPTMS mixture is added to
the sol. However, since eNH2 groups were not detected in the hy-
brids after drying, we can argue that the hypothesised reaction did
occur as the water was driven off during drying, producing hybrids
with covalent links between the components.
Acknowledgements
This work was supported by Fondazione Cariplo, grant n 2008-
3175 and Consorzio Italiano per le Biotecnologie (C.I.B). JRJ
acknowledges the EPSRC Challenging Engineering Scheme
(EP/I02086). EMV acknowledges the Natural Sciences and Engi-
neering Research Council of Canada (NSERC), Canadian Centennial
Scholarship Fund UK.
References
[1] L.L. Hench, J.M. Polak, Science 295 (2002) 1014e1017.
[2] J.R. Jones, L.M. Ehrenfried, L.L. Hench, Biomaterials 27 (2006) 964e973.
[3] B.M. Novak, Adv. Mater. 5 (1993) 422e433.
[4] G. Schottner, Chem. Mater. 13 (2001) 3422e3435.
[5] H.S. Mansur, R.L. Oréfice, A.P. Mansur, Polymer 45 (2004) 7193e7202.
[6] G. Andrade, E.F. Barbosa-Stancioli, A.A. Piscitelli Mansur, W.L. Vasconcelos,
H.S. Mansur, Biomed. Mater. 1 (2006) 221e234.
[7] Clément Sanchez, Philippe Belleville, Michael Popall, Lionel Nicole, Chem. Soc.
Rev. 40 (2011) 696e753.
[8] J.R. Jones, Acta Biomater. 9 (2013) 4457e4486.
[9] q. John, M. Ba1trukiewicz, P. Sobota, R. Brykner, q. Cwynar-Zaja˛ c, P. Dzie˛ giel,
Mater. Sci. Eng. C 32 (2012) 1849e1858. and references cited therein.
[10] E.M. Valliant, J.R. Jones, Soft Matter 7 (2011) 5083e5095.
[11] M. Vallet-Regi, Curr. Nanosci. 2 (2006) 179e189.
 L. Russo et al. / Materials Chemistry and Physics 140 (2013) 168e175
[12] D. Tian, P.H. Dubois, R. Jerome, J. Polym. Sci. Part A: Polym. Chem. 35 (1997)
2295e2309.
[13] S. Rhee, Biomaterials 25 (2004) 1167e1175.
[14] M.M. Pereira, J.R. Jones, R.L. Orefice, L.L. Hench, J. Mater. Sci. Mater. Med. 16
(2005) 1045e1050.
[15] L. Ren, K. Tsuru, S. Hayakawa, A. Osaka, Biomaterials 23 (2002) 4765e4773.
[16] Q. Chen, F. Miyajia, T. Kokubo, T. Nakamurab, J. Sol-Gel Sci. Technol. 19 (2000)
101e105.
[17] M. Catauro, M.G. Raucci, F. De Gaetano, A. Marotta, J. Mater. Sci. 38 (2003)
3097e3102.
[18] O. Mahony, O. Tsigkou, C. Ionescu, C. Minelli, R. Hanly, L. Ling, M.E. Smith,
M.M. Stevens, J.R. Jones, Adv. Funct. Mater. 20 (2010) 3835e3845.
[19] P. Judeinstein, C. Sanchez, J. Mater. Chem. 6 (1996) 511e525.
[20] J. Mackenzie, Q. Huang, T. Iwamoto, J. Sol-Gel Sci. Technol. 7 (1996) 151e161.
[21] W. Liu, X.H. Wu, H.B. Zhan, F.H. Yan, Mater. Sci. Eng. C: Mater. Biol. Appl. 32
(2012) 707e711.
[22] G. Poologasundarampillai, C. Ionescu, O. Tsigkou, M. Murugesan, R.G. Hill,
M.M. Stevens, J.V. Hanna, M.E. Smith, J.R. Jones, J. Mater. Chem. 20 (2010)
8952e8961.
[23] P. Innocenzi, C. Figus, T. Kidcho, M. Valentini, B. Alonso, M. Takahashi, Dalton
Trans. (2009) 9146e9152.
175
[24] M.A. Robertson, R.A. Rudkin, D. Parsonage, A. Atkinson, J. Sol-Gel Sci. Technol.
26 (2003) 291e295.
[25] P. Innocenzi, G. Brusatin, M. Guglielmi, R. Bertani, Chem. Mater. 11 (1999) 1672e1679.
[26] P. Innocenzi, G. Brusatin, F. Babonneau, Chem. Mater. 12 (2000) 3726e3732.
[27] M.Y. Koh, T. Miyazaki, C. Ohtsuki, J. Ceram. Soc. Jpn. 119 (2011) 387e392.
[28] L. Gabrielli, L. Russo, A. Poveda, J.R. Jones, F. Nicotra, J. Jiménez-Barbero,
L. Cipolla, Chem. Eur. J. (2012), http://dx.doi.org/10.1002/chem.201204326.
[29] P. Innocenzi, T. Kidchob, J. Sol-Gel Sci. Technol. 35 (2005) 225e235.
[30] D.S. Brauer, M. Mneimne, R.G. Hill, J. Non-Cryst. Solids 357 (2011) 3328e3333.
[31] L.L. Hench, J.K. West, Life Chem. Rep. 13 (1996) 187e241.
[32] R.A. Martin, S. Yue, J.V. Hanna, P.D. Lee, R.J. Newport, M.E. Smith, J.R. Jones,
Philos. Trans. R Soc. London A (2012) 1422e1443.
[33] S.R. Davis, A.R. Brough, A. Atkinson, J. Non-Cryst. Solids 315 (2003) 197e205.
[34] M.R. Gizdavic-Nikolaidis, Z.D. Zujovic, N.R. Edmonds, C.J. Bolt, A.J. Eastea,
J. Non-Cryst. Solids 353 (2007) 1598e1605.
[35] H.S. Costa, M.F. Rocha, G.I. Andra, E.F. Barbosa-Stancioli, M.M. Pereira,
R.L. Orefice, W.L. Vasconcelos, H.S. Mansur, J. Mater. Sci. 43 (2008) 494e502.
[36] John Coates, Encyclopedia of Analytical Chemistry, in: R.A. Meyers (Ed.),
Interpretation of Infrared Spectra, A Practical Approach, John Wiley & Sons
Ltd, Chichester, 2000, pp. 10815e10837.
[37] K.M. Delak, N. Sahai, J. Phys. Chem. B 110 (2006) 17819e17829.