HA/UHMWPE Nanocomposite Produced by Twin-Screw Extrusion FANG Liming1, LENG Yang*1, GAO Ping2 (1. Department of Mechanical Engineering, Hong Kong University of Science and Technology, Hong Kong 2. Department of Chemical Engineering, Hong Kong University of Science and Technology, Hong Kong)

Abstract: The HA/UHMWPE nanocomposite is compounded by twin-screw extrusion of the HA and UHMWPE powder mixture in paraffin oil and then compression molded to a sheet form. TGA measurement shows the HA weight loss after processing is about 1~2%. FTIR spectra indicate the paraffin oil residue is trivial and UHMWPE is not oxidized. SEM reveals the HA nano particles are homogeneously dispersed by twin-screw extrusion and the inter-particle spaces are penetrated with UHMWPE fibrils by swelling treatment. HRTEM image indicates the HA particles and UHMWPE are intimately contacted by mechanical interlocking. Compared with the unfilled UHMWPE, stiffness of the composite with the HA volume fraction 0.23 was significantly enhanced to 9 times without detriment of the yield strength and the ductility. Key words: Hydroxyapatite; UHMWPE; Composite; Twin-screw extrusion; Swelling

1 Introduction It has been a continuous effort to develop bio-analogue composites for medical applications, since Bonfield et al. introduced the hydroxyapatite (HA) reinforced high density polyethylene (HDPE) composite[1]. In our previous work, we have developed a HA/UHMWPE composite by wet ball milling and swelling treatment[2]. The composite exhibited an interpenetrated network structure with homogeneous HA dispersion, which was surrounded by a UHMWPE-rich phase due to the large UHMWPE/HA particle size ratio. In this study, we attempted to compound the HA and UHMWPE powder in paraffin oil through twin-screw extrusion and compression mold the swollen treated compound by hot press. Thus, the nano HA particles may disperse in the UHMWPE melt by twinscrew extrusion and the UHMWPE fibrils can penetrate in the inter-particle spaces and interlock with the HA particles.

2 Experiment HA/UHMWPE composite specimens with different HA volume fractions and pure UHMWPE specimens (Table 1) were processed by twin-screw extrusion the HA (National Engineering Research Center for Biomaterials, China) and UHMWPE (HiFax 1900, Basell Ltd, USA) powder mixture in paraffin oil (Merck, Germany) and then compression molding to a sheet form. Microstructure was examined by SEM and TEM. Mechanical properties were evaluated by tensile testing. Table 1 Sample list Material HA (vol%) HA0PE30Oil170* 0 HA5PE30Oil165 4.7 HA10PE30Oil160 9.0 HA15PE30Oil155 12.9 HA20PE30Oil150 16.5 HA25PE30Oil145 19.8 HA30PE30Oil140 22.8 *: number means weight in gram

Code PE C1 C2 C3 C4 C5 C6

3 Results 3.1 Chemical analysis TGA thermographs in Fig. 1 show that the HA weight loss after processing is about 1~2%. Three specimens are measured for each sample, and the standard derivation is less than 1%, which means the dispersion is very good. FTIR spectra of UHMWPE (a), paraffin oil (b), HA (c), HA/UHMWPE powder mixture (d) and the HA/UHMWPE composite (e) are shown in Fig. 2, which reveal the paraffin oil remaining in the composite is negligible at least it can not be detected in the

1

accuracy range of FTIR (Fig.2 inset). This is further confirmed by comparing the spectrum of the composite with those of the HA and UHMWPE powder mixture. No obvious carbonyl group bands appear in the range of 1650~1850 cm-1, which indicates that the UHMWPE matrix is not oxidized after processing at elevated temperature and high pressure in air.

a

Fig. 1 TGA thermograph

b

Fig. 2 FTIR spectra

Fig. 4 HRTEM image

Fig.3 SEM micrographs. a) ×250; b) ×30,000

Fig. 5 Modulus vs. HA volume fraction

3.2 Microstructure characterization Typical SEM micrographs of the HA/UHMWPE composite (VHA = 0.23) are shown in Fig. 3. Lower magnification (Fig. 3a) reveals a uniform global dispersion of HA in the UHMWPE matrix because the agglomerated HA powder is effectively broke down to its primary size by twin-screw extrusion the HA and UHMWPE powders in the paraffin oil. Higher magnification (Fig. 3b) shows that the UHMWPE fibrils penetrate in the inter-particle spaces and entangle the nano-sized HA

2

particles, which indicates the swelling treatment in paraffin oil enhances the chain mobility of UHMWPE remarkably. The macro-voids are observed in the composite (Fig. 3a) probably due to the deformation during cutting the sample. The HRTEM image (Fig. 4) taken around the edge of the HA particle reveals that the HA particles and the UHMWPE matrix intimately contact each other, and the fringe pattern of HA crystal plane (100) (d(100) = 8.2 Å) gradually disappears approaching the boundary, which indicates the interface is mechanically interlocked.

3.3 Mechanical properties Table 2 lists the mechanical properties of the UHMWPE and the HA/UHMWPE composite. Relative tensile modulus of the composite against the HA volume fraction is shown in Fig. 5. The modulus of the composite increases with the HA concentration gradually when the HA volume fraction is lower than 0.1, then it is enhanced almost linearly up to 9 times of that of pure UHMWPE when the HA volume fraction is 0.23. If the HA volume fraction is further increased to the level of cortical bone (0.4~0.5)[3], we may get a composite with stiffness similar to natural bone (10-30 GPa). Yield strength and ductility of composite are generally reduced with the amount of tough material available decreases as the interfacial fracture energy is usually much smaller than the polymer fracture energy. However, the yield strength of the HA/UHMWPE composite is almost the same as that of the unfilled UHMWPE, which means that there is certain interfacial adhesion between HA and UHMWPE. The HA particles show different effects on modulus and strength because the load transfer mechanism of stiffening is tensile while the strengthening is shear[4]. For the elongation at break, to our surprise, the results suggest that increasing the hydroxyapatite content of the composite to 0.23 without deleterious effects on the ductility of the composite (less than 25% decrease compared with pure UHMWPE). This phenomenon owes to the extremely ductile nature of UHMWPE, despite the fast interface failure. As the size of the filler is reduced to nanoscale, the composite becomes insensitive to flaws with elongation at break approaching the theoretical ductility of the matrix. Obviously, ductility of this material is sufficient for a bone substitute[3].

HA (vol%) 0 4.7 9.0 12.9 16.5 19.8 22.8

Table 2 Mechanical properties Modulus (MPa) Strength (MPa) 0.9±0.1 27.2±0.6 2.1±0.1 26.5±0.5 3.7±0.2 26.4±0.6 4.2±0.4 27.1±1.2 5.4±0.6 29.5±2.3 6.8±0.5 26.6±1.2 8.0±0.8 28.4±1.6

Ductility (%) 484±30 452±8 400±15 320±41 350±37 375±41 358±42

4 Discussion 4.1 Processing effects on the microstructure Compared with solid state compounding methods, twin-screw extrusion not only reduces the HA particle size but also melts the UHMWPE and then allows the HA particles disperse inside the melt. Therefore, a global homogeneous structure is obtained by twin-screw extrusion but a two-zone interconnected structure is produced by ball milling[2]. To break the HA agglomeration thoroughly, the mixture should be extruded several times to compensate the powder not at optimized position or orientation in the high stress zone of the extruder. In this study, the lower speed extrusion is used in order to break the agglomeration of HA powder and get a better dispersion, because the mixture has sufficient time to be mixed. The higher speed

3

extrusion is employed to generate higher shear forces and thus increase UHMWPE chain mobility to promote UHMWPE fibrils penetration in the spaces among HA particles. Swelling also plays a very important role in the processing because it greatly improves the chain mobility and reduces the chain entanglement density of UHMWPE. The higher chain mobility allows for better penetration of UHMWPE fibrils into the inter-spaces of HA particles. The UHMWPE fibrils may entangle the HA particles and transfer the load to the stiffer filler. The open spaces resulting from extraction of solvent are closed by compression molding at elevated temperature. Certainly, some submicron-sized pores may not be removed completely as indicated in Fig. 3b.

4.2 HA effects on the mechanical properties The elastic modulus of particulate reinforced polymer composites has been extensively studied assuming that the modulus of composites depends only on the volume fraction of filler and not the particle size, as reviewed by Ahmed and Jones[4]. All the models appear to satisfactorily furnish a phenomenological description of the experiment data. The modulus relationship with the HA volume fraction as shown in Fig. 5 is predicted using these models. However, it indicates that all models underestimate the experiment result. It is because that the effects of the filler size and interface interactions are not taken into account. However, it is highly unlikely that in practice, the smaller the particle size, the higher the specific surface area, thereby causing better packing efficiency and more filler-matrix interactions. Therefore, the reinforcement size and its dispersion and the way in which it is bound to matrix are the key factors in the analysis of the modulus of nanocomposites for a given volume fraction. The modulus of cortical bone is considerably higher than the models predicted because the mineral phase is staggered within the matrix in a complex hierarchical arrangement. The mechanical properties of most bio-analogue composites could not reach the level of natural bone with the same volume fraction of mineral phase, unless their structure can mimic natural bone. In our previous work, the HA particles are distributed in some ‘islands’, thus in response to the applied load the stress will be distributed unevenly. Microstructure of the HA/UHMWPE composite processed by twin-screw extrusion (Fig. 3b) exhibits a similar morphology with natural cortical bone, except the composite is at the sub-microstructure while the natural cortical bone is at the sub-nanostructure[3, 5].

5 Conclusion HA powder was dispersed homogeneously in the UHMWPE matrix by twin-screw extrusion. The chain mobility of UHMWPE was improved by swelling in the paraffin oil. The UHMWPE fibrils intimately contacted with the HA particles. Compared with the unfilled UHMWPE, stiffness of the composite was significantly enhanced to 9 times without detriment of the yield strength and the ductility.

Reference [1] W Bonfield, et al. Hydroxyapatite reinforced polyethylene - a mechanically compatible implant material for bone-replacement. Biomaterials 1981,2(3):185-186. [2] Fang LM, et al. Processing of hydroxyapatite reinforced ultrahigh molecular weight polyethylene for biomedical applications Biomaterials 2005,26(17):3471-3478. [3] Rho JY, et al. Mechanical properties and the hierarchical structure of bone Medical Engineering & Physics 1998,20(2):92-102. [4] Ahmed S, et al. A review of particulate reinforcement theories for polymer composites Journal of Materials Science 1990,25(12):4933-4942. [5] Weiner S, et al. The material bone: Structure mechanical function relations Annual Review of Materials Science 1998,28:271-298.

4

FANG Liming(方立明): Ph D candidate; E-mail: [email protected]; GAO Ping(高平):Assoc. Prof.; Ph D; E-mail: [email protected] Corresponding author: LENG Yang(冷扬): Assoc. Prof.; Ph D; E-mail: [email protected] *Funded by the High Impact Area at Hong Kong University of Science & Technology

5