Synthesis and characterization of porous hybrid biomaterials with improved mechanical properties

Introduction

One of the most important aspects in the development of biomaterials is their use as prosthesis or implants in living organisms, specifically for bone repair or bone ingrowth (traumatism, bone disease, etc.) applications.^1,2 Synthetic hydroxyapatite (Ca₅(PO₄)₃OH) (HAp) is the mineral most commonly used to produce biomaterials because it is chemically very similar to real bone. However, their acceptance by the living organism is based not only in the right chemistry, which is responsible to support cellular activity without eliciting an inappropriate host response, but also equally important is the morphology, which is responsible for providing the right conditions for vascularization. ^{3 – 5}

In recent years, polymer-ceramic composite materials with high content of HAp have been designed for biomedical applications (PA/HAp, PMMA/HAp, PLA/HAp, PE/HAp, etc.). ^{6 – 8} The design of these synthetic materials is based on the original constituents of bones: a structured hybrid composite made of organic (collagen, etc.) and mineral (HAp) constituents, which together confer to the final material high mechanical resistance.^9,10 When HAp is included in the design of biomaterials for clinical applications with concentrations around 65% wt, the composite is biocompatible and bioactive to initiate the osteogenesis. However, most of the studies on these materials showed either no increase or a slight decrease in compressive strength of the polymer materials when HAp was incorporated.^11,12

The bioceramics are widely used in bone grafting and dental devices because HAp has the ability to favor the presence of osteoblasts required for bone tissue engineering.^13,14 The chemical properties of the ceramic surfaces determine their compatibility with the resin and the performance of the whole material.^15,16 Other ceramic materials are biocompatible, osteoconductive, and bioabsorbable and are commonly used as bone repair, like calcium sulfate (CS), which is used to fill bone defects and maxillary sinus lifts because it is well tolerated by the host, is completely re-absorbed within 3 months, and does not interfere with bone healing. However, the HAp–polymer composites combine the excellent biocompatibility, bioactivity, and osteoconductivity of HAp with the important mechanical properties of polymeric materials, producing composites with improved properties. ^{17 – 19} It has been reported that high-purity high-density (>3.9 g/cm ³ ) alumina was the first bioceramics used for implants, specifically in load-bearing hip prostheses and dental implants because of its excellent corrosion and wearing resistance. Alumina is considered as bio-inert material, safe to use in the human body, but it is not bonded to living tissue when implanted.^20,21

Recently, the fabrication of bioceramics with controlled morphology (porosity) has attracted considerable attention because organic tissue can grow in the interior of the interconnected pores allowing the vascularization of the implant and enhancing the implant-tissue interaction. The colonization of the pores takes place when the pore size is in the range from 100 to 300 µm and the pore volume fraction higher than 50% vol. ^{22 – 24}

In the last decades, an increasing demand for high-durability biomaterials has been observed; then, it is important to provide biomaterials with the right properties to stay for longer times into the living organism. It is known that the addition of hard ceramic particles improves the mechanical and tribological properties of polymers: the hardness and size of ceramic particles strongly affect these properties in the final hybrid material.^25,26 In this case, the mechanical properties and the abrasion and the scratching resistance were increased by the addition of two different types of particles with different sizes: alumina microparticles and silica nanoparticles. ²⁷ The use of ceramic particles with different sizes allows the filling, with nanoparticles, of the interstitial space left by the microparticles; the use of micro- and nanoparticles produces a reinforcement of the composite that renders in materials with improved mechanical and tribological properties.^28,29

The mixture of polyisocyanate with the resin and the ceramic particles is a reactive system because the polyisocyanate reacts with all OH groups present in the system. If the silica nanoparticles are added first, because their surface area is considerably larger with respect to the alumina microparticles, the number of OH groups that can react with the isocyanate is also larger, producing a highly cross-linked material in a short time (i.e., a fast cure process); this increases significantly, and in a short time, the system viscosity reducing the possibility to obtain a homogeneous mixture; additionally, in this case, the amount of ceramic particles that can be introduced into the polymer is low respect to the case where the alumina microparticles are introduced first.^30,31 In this last case, the number of OH groups that can react with the isocyanate is considerably lower, allowing a slow curing process with an associated slow increment in viscosity that allows a good homogenization of the composite. When the silica nanoparticles are added after the alumina microparticles, the nanoparticles fill the interstitial spaces left by the largest particles, producing an increment in the mass density that renders in an improvement in the mechanical and tribological properties. In this case, the amount of ceramic particles that can be introduced into the polymer is larger with respect to the former case. ²⁷ In this work, new ceramic-polymers porous hybrid materials were synthesized with the appropriated morphology to be used as prosthesis or implants in living organism. The materials were added with two different types of ceramic particles of different sizes to improve the mechanical and tribological properties of the final material.

Experimental

Samples preparation

The samples were prepared using solvent-free mono-component polyester polyurethane containing a malonate-blocked polyisocyanate and three different types of ceramic particles with different sizes. A total of 8 g of the mono-component polyester polyurethane were mixed with the corresponding concentration of ceramic particles (Table 1). Because the ceramic particles have different sizes, the largest particles (HAp) were added first, followed by the alumina microparticles (3.2 microns; Cabot, USA) and finally by the silica nanoparticles (16 nm; Degussa, Ger). Each type of ceramic particles was well incorporated before the addition of the next type. The incorporation was carried out with strong and high-shear rate agitation until a smooth and homogeneous paste was obtained. The addition of the ceramic particles unstabilizes the polymer emulsion and the chemical reaction between the ciano groups of polyisocianate and the hydroxyls groups of the resin and ceramic particles starts releasing part of the malonate blocking structure as CO₂ that generates the pores into the material; this reaction is temperature dependent. A change in the temperature of the thermal treatment allows to control the pore size and the pore volume fraction.

OPEN IN VIEWER

Table 1.

Composition of all prepared materials

Sample	PU	HAp	Al2O3	SiO2
50Si-50HAp	8.0 g	3.8 g	0 g	3.8 g
70A-30HAp	8.0 g	2.3 g	5.3 g	0 g
80A-20HAp	8.0 g	1.5 g	6.1 g	0 g

Samples of different sizes and geometries were prepared according to the corresponding test. A sample of each formulation was kept into the mould for 24 h at room temperature to finish the curing reaction. Samples of each formulation were thermally treated at different temperatures and times: 50°C (12 h), 80°C (8 h), 110°C (2 h), and 150°C (1 h); three samples were prepared for each composition and for each temperature.

Samples characterization

The size of the ceramic particles was determined using a dynamic light scattering (DLS) apparatus Brookhaven Instruments Corp. model BI-APD equipped with a He-Ne laser at 632.8 nm and a digital correlator. The average particle size of the ceramic particles was 16 nm for silica and 3.2 microns for alumina.

The mass densities were determined weighting the samples in an analytical balance with a resolution of 10⁻⁵ g and measuring the sizes using a micrometer with resolution of 0.01 mm and an accuracy of ±0.001 mm. The pore size was obtained using the scanning electron microscopy (SEM) images and the pore volume fraction was determined using the formula reported elsewhere ³³

(1)

where ϕ_p, ρ_p, and ρ_s are the pore volume fraction, the density of the porous material, and the density of the non-porous material (prepared using no-blocking isocyanate).

The mechanical tests were performed according to the norm ASTM D-695-02 a in an Adamel Lhomargy machine model DY.22 in compression mode with a compression rate of (1.3 ± 0.3) mm/min. The cell load was of 5000 N with a resolution of 0.1 N. For this test, cylindrical samples were prepared of each formulation using a Teflon mould of 20 mm high and 10 mm diameter.

The wearing resistance was determined using the Taber Method according to the norm ASTM-D-1242-95a; ³⁴ the weight lost was obtained by sanding the flat surface of the samples with a F-400 Fandeli sandpaper mounted on a steel plate rotating at 76 rpm in dry conditions; the samples, with dimensions 2.0 × 1.0 × 0.5 cm, were loaded with a weight of 20 g. Five experiments were performed for each sample, all at room temperature. The weight lost was determined every 20 s with an accuracy of ±1 × 10⁻⁵ g. After each sanding procedure and before the weight determination, the scratched surface was cleaned with a dry soft cloth wipe to remove the dust and the sand paper was also cleaned with a soft brush. The abrasion resistance, ³⁴ defined as the time required to remove a unit mass of the material, corresponds to the inverse of the weight lost.

The water absorption (WA) was determined immersing completely the samples, of known weight, in distilled water for 21 days and determining the final weight; after the immersion time, the samples were superficially dried with a tissue paper and immediately weighted. The WA was obtained according to the expression

(2)

where W_i and W_f are the initial and final weights.

The SEM was performed in a JEOL JSM-6060 at 20 kV in secondary electron mode with different magnifications; the samples were frozen in liquid N₂, broken, and the new exposed surface covered with a gold film. The average pore size was determined using the SEM images.

Results

The mass densities of all prepared materials are reported in Figure 1. In all cases, the density decreases practically linear with temperature, reducing from near 1 g/mL to 0.4 g/mL when the temperature was varied from room temperature to 150°C. The pore size and the pore volume fraction are reported in Figures 2 and 3. As a consequence of the reduction in density, the pore size and the pore volume fraction increase with temperature; the pore volume fraction is directly related to the density, and then grows also practically linear with temperature with slopes in the range from 0.3 to 0.45 (°C)⁻¹.

OPEN IN VIEWER

Figure 1.

Mass density as a function of temperature.

OPEN IN VIEWER

Figure 2.

Average pore size obtained from the scanning electron microscopy (SEM) images.

OPEN IN VIEWER

Figure 3.

Pore volume fraction obtained using Equation (1).

Figure 4(a) through (c) shows the results of the mechanical properties. In Figure 4(a) is reported the Young modulus as a function of the temperature while in Figure 4(b) and (c) the force and the deformation to rupture, respectively. Figure 5 shows the abrasion resistance (the inverse of the rate the weigh is lost) as a function of temperature; here it is possible to see that for all samples the highest abrasion resistance was obtained for samples treated at 110°C. The weight percentage of the water absorption is reported in Figure 6 where, as expected, this quantity increases with the porosity and consequently with temperature. Typical SEM images of the samples are shown in Figure 7(a) through (d).

OPEN IN VIEWER

Figure 4.

Mechanical properties: (a) Young modulus, (b) force to rupture, and (c) deformation to rupture.

OPEN IN VIEWER

Figure 5.

Abrasion resistance as a function of temperature.

OPEN IN VIEWER

Figure 6.

Percentage of water absorption obtained using Equation (2).

OPEN IN VIEWER

Figure 7.

A typical series of scanning electron microscopy (SEM) images for the sample 80A-20HAp at different temperatures: (a) 150°C; (b) 110°C; (c) 80°C; (d) 50°C.

Discussion

The reduction in density when the temperature was increased (Figure 1) modifies all properties of the samples: the morphology, the mechanical and the abrasion properties, and the water absorption. The reduction in density (from 1.0 to 0.4 g/mL) follows a linear dependence with temperature where the slopes are in the range from 0.003 to 0.004 g/mL°C depending on the type and concentration of the ceramic particles. It is important to note that, for a given temperature, the higher densities were obtained for the sample containing silica nanoparticles. The chemical reaction between the blocked isocyanate and the hydroxyls groups of the resin and the ceramic particles is responsible for the CO₂ production, that is, the pore generator. This reaction is temperature dependent as evidenced by Figures 2 and 3, where the pore size and the pore volume fraction are reported as a function of temperature. The pore size (Figure 2) increases with temperature because this accelerates the chemical reaction producing more and more rapidly CO₂ gas into the material; additionally, the pressure of the CO₂ is also temperature dependent, increasing linearly with temperature (ideal gas law). These two effects reinforce each other to increase the pore size and the pore volume fraction with temperature; this is in full agreement with the results shown in Figures 2 and 3. These pore size profiles show the presence of two regimes: the first one at low temperature where the pore size grows slowly with the temperature, and the second one starting between 80 and 110°C and characterized by a faster increment in the pore size with temperature. The change in regime (inflection) happens at temperatures between 80 and 110°C depending on the type of the ceramic particles: for alumina-containing particles the transition happens at a lower temperature (80°C), while for the silica-based sample the inflection happens at higher temperature (110°C).

The pore volume fraction was calculated using Equation (1) and the results are reported in Figure 3. As mentioned, a non-porous reference sample was prepared using a non-blocked isocyanate in order to avoid the CO₂ generation; for this material the density was 1.04 g/mL. Because the pore volume fraction is directly related to the mass density, this increases with the temperature in a practically linear way. At low temperature (bellow 50°C), the pore volume fraction is below the percolation threshold and consequently the pores are not interconnected.

As mentioned, porous materials possess generally weak mechanical properties essentially because the pores act as stress concentrators on a thin wall favoring the presence of cracks and fissures and diminishing the mechanical properties. This behavior was observed in Figure 4(a) where the Young modulus is reported as a function of temperature; here it is possible to see that the Young modulus was reduced with the temperature for low temperatures; however, at higher temperatures, this reduces slowly, and in some cases increases, having values larger than expected. This is because at higher temperatures there are two effects that are in opposition: on one side the weakness of the material produced by the increment in porosity and on the other the increment in the chemical links between the ceramic particles and the polymeric resin which improves the mechanical properties. Due to this, samples treated at higher temperatures showed a reduction in the rate of diminishing (sample 80A-20HAp) or even an increment (samples 70A-30HAp and 50Si-50HAp) in the Young modulus for temperatures higher than 110°C. The profile for the sample containing silica is different possibly due to the densification process (Figure 1) produced by the filling of the interstitial space left by the HAp micro-particles with the silica nanoparticles. As mentioned, in all cases the Young modulus have higher and similar values at 150°C.

It is also expected a reduction in the force to rupture for samples synthesized at higher temperatures which possess higher porosities. However, Figure 4(b) shows that this quantity is less sensitive to the thermal treatment, having similar values in the whole range of temperatures. This means that the increment in porosity produced by the thermal treatment is compensated by the increment in the chemical links in the ceramic-resin interface. Something similar happens with the deformation to rupture (Figure 4(c)): the temperature has not a strong effect in the deformation to rupture for all samples, meaning that this is practically controlled by the polymeric phase and the polymer-ceramic interface: at 150°C the deformation to rupture is the same for all samples.

Figure 5 shows the abrasion resistance as a function of temperature. As can be noticed, the samples treated at 110°C showed the highest abrasion resistance, followed by a reduction at 150°C. This behavior is consistent with the explanation provided before the weakening produced by the increment in porosity is partially compensated by an increment in the chemical links between the ceramic particles and the polymeric resin.

The morphology of the materials has an important effect on the water absorption; this is reported in Figure 6. As can be noticed, samples treated at higher temperatures absorb more water; this is expected because when the porosity is increased, more water can be trapped into the pores in the interior of the material; additionally, at high temperature (i.e. at higher porosity) the pores are percolated allowing the water penetrates deeper in the interior of the material, increasing its absorption. As can be noticed at 150°C, all samples have practically the same amount of absorbed water (between 32 and 34%). As before, there are two regimes in the water-absorbed profiles: for temperatures lower than 110°C a considerable fraction of pores are blind reducing the amount of water that can be absorbed by the sample, on the other hand at temperatures higher than 110°C the pore volume fraction exceed the percolation threshold allowing the water to be absorbed deeply into the sample.

A typical series of SEM images for the sample containing 80% of alumina and 20% HAp is reported in Figure 7(a) through (d): (a) 150°C; (b) 110°C; (c) 80°C; (d) 50°C. Images for other samples, not reported here, are similar.

Conclusions

It was possible to produce hybrid porous materials with good mechanical and tribological properties to be used as biomaterial for bone substitution. The porosity was controlled with the temperature of the thermal treatment: increasing the temperature from ambient to 150°C allows to increase the porosity reducing the density from 1 to 0.4 g/mL. The addition of ceramic particles of different sizes allows to produce highly porous biomaterials with improved mechanical and tribological properties. When the temperature of the thermal treatment was 110°C or higher, it was possible to obtain porosities larger than 40% (for 70A-30HAp and 50Si-50HAp), allowing the vascularization through biomaterial without a significant reduction in the mechanical properties.