a novel method to prepare poly(dimetil siloxane)/calcium

advertisement
A NOVEL METHOD TO PREPARE POLY(DIMETIL SILOXANE)/CALCIUM
PHOSPHATE COMPOSITES USING THE BIOMIMETIC APPROACH
Bareiro, O.(1); Santos, L. A.(1)
(1)
Laboratorio de Biomateriais, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brasil
oscarbafer@hotmail.com, luis.santos@ufrgs.br
Abstract: The main objective of this study was to develop a method to prepare poly(dimetil
siloxane) (PDMS) / calcium phosphates composites by using the biomimetic technique. This
process involved the precipitation of calcium phosphate particles within the PDMS matrix
during the composite fabrication. The chemical composition of the filler phase was studied by
means of X-ray diffraction (XRD) and energy dispersive X-ray spectroscopy (EDS). Scattering
electron microscopy (SEM) micrographs showed tiny particles (200 nm) of calcium phosphate
evenly distributed and dispersed on the composite surface. The influence of the filler phase on
the composite thermodynamics parameters of cross-linking, static contact angle and roughness
was investigated. It was determined that the introduction of calcium phosphate into the PDMS
matrix changed the enthalpy of cross-linking, data from the calorimetric analysis showed that
the cross-linking enthalpy of the composites increases in comparison to the unfilled PDMS.
Results from the static contact angle assays indicated a reduction in the PDMS hidrofobicity and
an improvement in the roughness (Ra), once compared to the unfilled PDMS.
Keywords: biomimetic method, calcium phosphate, nanoestructured composite, reinforcement,
silicone elastomer.
1. INTRODUCTION
Silicone rubber (SR) has been widely used in clinic as implant for a long time. It is well known
that SR is of good biocompatibility and physiological inertness. Its well workable property
promises a great convenience in clinical applications. However, SR is not able to unite
organically with tissues owing to its inertness. In some cases inflammation and foreign-body
reaction occur after the implantation [1]. In recent years, a number of studies have focused on
improving the mechanical, physicochemical and biological properties of polymers such as SR
and chitosan scaffolds through incorporation of bioactive inorganic substances, such as calcium
phosphate or hydroxyapatite (HAp) [Ca10(PO4)6(OH)2], especially to achieve porous tissue
ingrowth and high mechanical strength [2].
The resulting SR/HAp composite can be viewed as representing a new class of nanostructured
biomaterial. The composite has the potential to exhibit excellent physical and biological
properties, with the SR providing the desired fine mechanical properties and HAp acting
synergistically to promote bioactivity. The compounding of SR with HAp will provide bioactive
sites that are bioresorbable and favorable to tissue ingrowth [2].
In nature, the nucleation and growth of mineralized materials are often controlled by organic
macromolecules such as proteins and polysaccharides. Bone and teeth consist of a small amount
of organic matrix which manipulates the formation of apatite into distinct microstructures
suitable for the mechanical forces they encounter in vivo [3]. A new development in biomaterials
is the biomimetic synthesis of calcium phosphate in polymer matrices to produce composites that
can initiate osteogenesis when implanted in bony sites [4]. Polymers with distinct molecular
organizations may be used as a template to control the geometry of the apatite to mimic that
found in bone [3].
This method for introducing reinforcing particles has a number of advantages over the
conventional approach in which separately-prepared filler particles are blended, with difficulty,
into the uncross-linked elastomer before its vulcanization [5, 6]. Because of the nature of the in
situ precipitation, the particles are well dispersed and are essentially unagglomerated. The
mechanism for their growth seems to involve simple homogeneous nucleation, and since the
particles are separated by polymer, they do not have the opportunity to coalesce [7]. The main
objective of this study was to develop a method to prepare poly(dimetil siloxane) (PDMS) /
calcium phosphates composites by using the biomimetic technique.
2. MATERIALS AND METHODS
In order to produce the PDMS/calcium phosphates composite, 10 - 30 %(v/v) of Ca(OH)2 (95%
Synth) was incorporated to PDMS matrix (NE-140, Wenda Co.) in a open two-roll mixer. After
mixing, H3PO4 was added to obtain a Ca/P ratio of 1/1, the system was once again mixed. 0.8
%(wt/wt) of the cross-linking agent, dicumile peroxide (98% Aldrich) was added to the system.
The pH values of a aqueous solution containing 0.5 g of the composites were registered. Then,
the composite were placed in metallic molds and pressed in order to obtain 3 mm thick samples
and was cross-linked at 185 °C for 35 min.
The phase composition of the obtained composite was characterized by X-ray diffraction (XRD)
(Phillips X'Pert MPD, Kα = 1.5418 Å) and its surface and fracture morphology was studied by
scanning electron microscopy (SEM) (JEOL JSM 5800) coupled with a energy disperse
spectroscopy (EDS) equipment. In order to determine thermodynamic parameters of the cure,
samples of the composite were analyzed by differential scanning calorimetric (DSC Q20 V24.2
Build 107). Measurements were made in nitrogen atmosphere, with 5 mg samples, the heating
rate was of 10 °C/min. In order to evaluate the filler influence on the PDMS surface energy,
static contact angles were measured using water as liquid. The reported angles consist of an
average of 7 independent measurements, which were performed at room temperature using the
sessile drop method. The values of surface free energy were obtained by Neumann’s method [8].
The contact angle measurements were made using the ImageJ software. The surface roughness
(Ra) of the composite was assessed using a surface roughness measurer (Mitutoyo SJ-400)
following the ASTM D2240 standard, the reported values consist of 5 independent measures
maid at room temperature.
3. RESULTS AND DISCUSSION
DRX difractograms of the PDMS/20%(v/v)Ca(OH)2 composite, corresponding to the elaboration
process, are shown in Fig. 1. Difractogram Fig. 1(a) is related to the elastomeric matrix filled
with Ca(OH)2 (JCPDS 01-076-0570) before the cross-linking, Fig. 1(b) corresponds to the
addition of H3PO4, which led to the precipitation of dibasic calcium phosphate dihydrate (DCPD)
(JCPDS 02-0085) with some residual Ca(OH)2, this reaction could be express as follows;
Ca(OH)2(sol) + H3PO4(aq)  CaHPO4.2H2O(sol) (DCPD) + Ca(OH)2’(sol)
Eq. 1
As shown in difractogram Fig. 1(c), after the composite cross-linking at 185 °C, DCPD
dehydrated leading to the precipitation of dibasic calcium phosphate anhydrous (DCPA) with some
residual Ca(OH)2, such reaction could be represented as follows;
CaHPO4.2H2O(sol) (DCPD) + Ca(OH)2’(sol)  CaHPO4(sol) (DCPA) + Ca(OH)2’(sol) + 2H2O(gas) Eq. 2
.
Δ
Figure 1. XDR difractograms obtained during the elaboration process of the PDMS/20%(v/v)Ca(OH) 2 composite.
(a) incorporation of Ca(OH)2 to the elastomeric matrix, (b) addition of H 3PO4 and precipitation of dibasic calcium
phosphate dihydrate (DCPD) with remaining Ca(OH)2, (c) composite after cross-linking, showing dibasic calcium
phosphate anhydrous (DCPA) with remaining Ca(OH)2.
Fig. 2 shows SEM fracture micrographs of the PDMS/20%(v/v)Ca(OH)2 composite before the
cross-linking, whilst Fig. 3 shows SEM fracture micrographs of the PDMS/20%(v/v)Ca(OH)2
cross-linked composite after the addition of H3PO4. In both cases, EDS chemical analysis spectra
are also shown. In the case of the cross-linked composite, values of the Ca/P ratio are indicated
in Fig. 3(d).
Figure 2. (a) SEM fracture surface micrograph of PDMS filled with 20 %(v/v) Ca(OH)2, (b) punctual analysis
spectra and surface analysis spectra of PDMS, corresponding to the (c) Ca and (d) Si elements.
Figure 3. (a) SEM fracture surface micrograph of the PDMS/20%(v/v)Ca(OH)2 composite, after the incorporation of
H3PO4, (b) punctual analysis spectra and surface analysis spectra of the composite, corresponding to the (c) Ca, (d) P
and (e) Si elements.
Fig. 2(a) shows that the dispersion and distribution state of Ca(OH)2 particles was homogeneous
in the elastomeric matrix; at the same time as Fig. 3(a) shows that the dispersion and distribution
state of the precipitated DCPA particles was as well homogeneous.
Fig. 3(d) shows Ca/P ratio values of the calcium phosphate phase, this values were slightly
greater that the Ca/P ratio corresponding to pure DCPA (Ca/P = 1,0). As observed in
difractogram Fig. 1(c), there is certain amount of residual Ca(OH)2 along with the precipitated
DCPA. Such difference between the Ca/P ratio values could be due to the remaining Ca(OH)2
presented in the particles surface, leading to a increase in the Ca/P ratio, in that particular
position. In order to promote the complete transformation of Ca(OH)2 into DCPA, during the
composite elaboration process, 0.94 mol/g of a aqueous NaOH (5 M) solution was added to
composite, this with the purpose of solubilizing the remaining Ca(OH)2 to allow its
transformation into DCPA.
Fig. 4 shows the difractograms of the composites treated with the NaOH solution, Tab. 1
presents the pH values of those composites. It was possible to indentify the phase DCPA, peaks
corresponding to Ca(OH)2 were not detected, due to its complete solubilization and
transformation into DCPA. In difractogram Fig. 4(e), corresponding to the
PDMS/30%(v/v)Ca(OH)2 composite, the main peaks related to hidroxyapatyte (JCPDS 9-0432)
were detected. According to Tab. 1, only this sample reached the necessary pH value to
precipitate this phase. Meanwhile, all the pH values were in the recommended range for
implantable materials [9]. Fig. 5 shows SEM external micrographs of the composites surface
morphology.
Figure 4. XDR difractograms of the PDMS/Ca(OH)2 composites treated with a NaOH solution, obtained with (a) 10
%, (b) 15 %, (c) 20 %, (d) 25 % and (e) 30 %(v/v) of Ca(OH) 2.
Table 1. pH values of the composites treated with the NaOH solution.
Figure 5. SEM micrographs showing the external surface morphology of the PDMS/Ca(OH)2 composites, obtained
with (a), (b) 10 %; (c), (d) 20 % and (e), (f) 30 %(v/v) of Ca(OH) 2.
Based in the results shown in Fig. 5 it is possible to observe, regardless the composite
composition, particles with a average size of 200 nm in a homogenous state of dispersion and
distribution. The greater the volume fraction of Ca(OH)2 added, the grater the size of the
agglomerates formed by this particles.
Fig. 6 shows DSC termograms of the composites cross-linking. It is possible to observe that all
DSC termograms were characterized by a similar profile, showing a broad peak at 100 – 125 °C,
related to the DCPD dehydratation, leading to the precipitation of DCPA; and a well defined
peak with a maximum at ~ 185 °C, corresponding to the cross-linking of the elastomeric matrix.
Figure 6. DSC termograms of the PDMS/Ca(OH)2 composites.
Tab. 2 shows some thermodynamics features calculated from the termograms. From these results
it’s possible to observe that the incorporation of the filler led to lower values of cross-linking
enthalpy (ΔH) in the composites. This due to the superior thermal conductivity of calcium
phosphates, once compared to pure silicone; which improved heat propagation, enhancing the
cross-linking reaction. Furthermore, the filler incorporation decreased the energy needed to the
composite cross-linking reaction. The greater the volume fraction of Ca(OH)2 added, the grater
the composites ΔH values.
Meanwhile, the filler incorporation led to higher Tbeg values, this is, the required temperature to
initiate the cross-linking reaction shifted to higher temperatures, this due to the inferior calorific
capacity of the filler, leading to greater values of Ti for the same amount of cross-linking energy.
Tmax and Tf did not present any significant modification.
Table 2. Values of cross-linking thermodynamics features of the PDMS/Ca(OH) 2 composites.
Tbeg: temperature of the beginning of cross-linking process; Tmax: temperature of the maximal rate of cross-linking
process, i.e. maximum of the peak; Tend: temperature of the end of the cross-linking reaction.
Fig. 7 shows static contact angle values of the composite. Based on this results it is possible to
elucidate that the filler incorporation led to lower contact angle values in the composites, once
compared to pure silicone. Pure silicone presented a static contact angle of 115,71º ± 0,90, whilst
the PDMS/30%(v/v)Ca(OH)2 composite showed a static contact angle of 100,60º ± 2,30, which
represents a 15,02 % ± 0,16 reduction.
Figure 7. Static contact angle values of the PDMS/Ca(OH)2 composites.
The calculated surface energy values from the static contact angle measurements are presented in
Fig. 8. As result from those measurements, only the composites PDMS/25-30%(v/v)Ca(OH)2
were influenced by the fillers introduction. The addition of such Ca(OH)2 volume fractions led to
lower surface energy values in the composite, once compare to pure silicone.
Figure 8. Surface energy values of the PDMS/Ca(OH)2 composites.
The results of the surface roughness (Ra) measurements are shown in Fig. 9. From these results it
is possible to elucidate that the introduction of Ca(OH)2 led to higher values of surface
roughness. Adding 15 – 25 %(v/v) of Ca(OH)2 a roughly constant value of 7.0 μm was reached,
whilst the PDMS/30%(v/v)Ca(OH)2 composite showed a surface roughness of 10,9 ± 1,5 μm.
Figure 9. Surface roughness (Ra) of the PDMS/Ca(OH)2 composites.
4. CONCLUSION
A new method was developed to elaborate PDMS/calcium phosphates composites by using the
biomimetic technique. This method produced well dispersed and distributed DCPD nanometric
particles within the PDMS matrix. The residual Ca(OH)2 has transformed into DCPD by the
addition of a suitable volume of NaOH solution during the composite processing. This process
led to composites with pH values within the recommended range for implantable materials.
The introduction of the fillers led to lower values of cross-linking energy to the PDMS and to
lower values of static contact angle, and subsequently to higher values of surface energy. Higher
values of surface roughness also were achieved.
REFERENCES
[1] ABBASI, F.; MIRZADEH, H.; KATBAB, A. A. Modification of polysiloxane polymers for
biomedical applications: a review, Polymer International, v. 50, p. 1279-1287, 2001.
[2] THEIN-HAN, W. W.; SHAH, J.; MISRA, R. D. K. Superior in vitro biological response
and mechanical properties of an implantable nanostructured biomaterial: Nanohydroxyapatite–
silicone rubber composite, Acta Biomaterialia, v. 5, p. 2668-2679, 2009.
[3] STUPP, S.I.; BRAUN, P.V. Molecular manipulation of microstructures: biomaterials,
ceramics, and semiconductors. Science, v. 277, p. 1242-1248, 1997.
[4] KOKUBO, T.; KIM, H.M.; KAWASHITA, M. Novel bioactive materials with different
mechanical properties. Biomaterials, v. 24, p. 2161-2175, 2003.
[5] RIGBI, Z. Reinforement of rubber by carbon black. Adv Polym Sci, v. 26, p. 21-68, 1980.
[6] MEDALIA, A.I.; KRAUS, G. Reinforcement of elastomers by particulate fillers. In: Mark
JE, Erman B, Eirich FR, editors. Science and technology.
[7] PAUL, D. R.; MARK, J. E. Fillers for polysiloxane (“silicone”) elastomers, Progress in
Polymer Science, v. 35, p. 893-901, 2010.
[8] NEUMANN, A. W.; GOOD, J. R.; HOPE, C. J.; SEJPAL, M. An equation-of-state approach
to determine surface tensions of low-energy solids from contact angles, J. Colloid Interface Sci.
v. 49(2), p. 291, 1974.
[9] DRIESSENS, F. C. M. In: Bioceramics Vol 8 - 4° Euro Ceramics, 1998, Bologna/Italia.
Anais. Faenza: Faenza Editrice, p. 77-83, 1997.
Download