Influence of biological lubricant on the morphology of UHMWPE wear particles generated with microfabricated surface textures

Materials Chemistry and Physics 95 (2006) 280–288

Influence of biological lubricant on the morphology of UHMWPE wear particles generated with microfabricated surface textures Hsu-Wei Fang a,∗ , Yu-Chih Su a , Chun-Hsiung Huang b , Charng-Bin Yang c a

Department of Chemical Engineering and Biotechnology & Institute of Biotechnology, National Taipei University of Technology, Taipei, Taiwan b Department of Orthopaedics, Mckay Memorial Hospital, Taipei, Taiwan c Department of Orthopaedics, Taipei Municipal Chung-Hsin Hospital, Taipei, Taiwan Received 12 January 2005; received in revised form 26 May 2005; accepted 6 June 2005

Abstract Immunological response induced by wear particles of ultra-high molecular weight polyethylene (UHMWPE) has been recognized as the major factor causing the failure of total joint replacements. A previous study has applied surface texturing techniques to generate UHMWPE wear particles with specific size and shape to study the effects of particle size and shape on osteolysis. In the present study, the effects of biological lubricants on the morphology of UHMWPE wear particles generated with microfabricated surface textures were investigated. It was observed that UHMWPE wear particles generated in bovine serum are smaller and thinner than the particles generated in water. The reason may be due to the reduction of the friction between UHMWPE and the surface feature under serum lubricating condition. It means a smaller material resistant force to overcome during the surface-feature sliding process and leads to a smaller lateral displacement of a micro-cutting process. Thus a larger aspect ratio (or a smaller particle width) was observed for the particles generated in serum. Compared to original UHMWPE, highly cross-linked UHMWPE has better wear resistance and generates smaller wear particles under the articulation with microfabricated surface textures in a biological environment. The potential application is to apply surface textures on the articulating surface of joint implant in order to control the size and shape of UHMWPE wear particles. While maintaining a low wear rate of UHMWPE parts, further reduction of the most “toxic” particles released into human body shall prevent particles induced osteolysis. © 2005 Elsevier B.V. All rights reserved. Keywords: UHMWPE; Wear particles; Surface texture; Lubricant; Microfabrication; Serum

1. Introduction Aseptic loosening from osteolysis has been recognized as the cause for joint implant failures [1]. UHMWPE wear particles are generated and released in surrounding tissues from the articulation between the polymer and metal parts of joint replacements. Phagocytosis of particles by macrophage cells results in the subsequent immunological responses and leads to the osteolysis [2,3]. Retrieval studies have shown that the size of the UHMWPE wear particles ranges from sub-micrometer to millimeters and a variety of shapes (granules, beads, fibrils, and shreds). The effects of size and shape of UHMWPE wear particles on the bioactivity are critical

∗

Corresponding author. Tel.: +886 930380200; fax: +886 227418575. E-mail address: [email protected] (H.-W. Fang).

0254-0584/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.matchemphys.2005.06.020

to the failure of total joint implants. For this reason, bioresponse studies of wear particles have been pursued by many researchers [4–13]. However, only spherical or roundlike UHMWPE particles were tested due to the difficulty in obtaining polyethylene particles with different sizes and shapes. For this purpose, one needs aseptic particles, sufficient quantities of particles, and various sizes and shapes of UHMWPE particles for further study. Fang et al. [14–16] have successfully applied surface texturing techniques to prepare narrowly distributed UHMWPE particles with controlled sizes and shapes. The principles of surface texture design to achieve wear particle morphology control in water have been developed. This accomplishment enables the bioactivity evaluation of different sizes and shapes of UHMWPE wear particles [17]. Ren et al. adopted the murine air pouch animal model to evaluate the immunological

H.-W. Fang et al. / Materials Chemistry and Physics 95 (2006) 280–288

responses caused by various particle sizes and shapes [18]. It was observed that elongated particles induce stronger biological response indicated by the pouch thickness and released cytokines. Smaller particles within the phagocytosable range stimulate higher immunological response [16–18]. A biophysical analysis of phagocytosis process pointed out that the phagocytosis capacity of the macrophage cells is a function of particle size and shape [18]. Compared to spherical particle with the same volume, elongated particles have larger surface area and occupy more phagocytosis capacity. It provides the explanation of particle shape effect on bioactivity. Once the effects of particle size and shape on biological response are understood, it is possible to enhance the life of total joint replacement by reducing the production of the most toxic particle populations in terms of size and shape. Our idea is to apply surface textures on the articulating surface of joint implant in order to control the size and shape of UHMWPE wear particles. While maintaining a low wear rate of UHMWPE parts, further reduction of the most “toxic” particles released into human body shall prevent particles induced osteolysis. Our previous studies have revealed the feasibility to control the UHMWPE wear particle morphology in water. In the present study, we further explore the relationship between the UHMWPE particle morphology and surface texture dimensions under a biological lubricating condition. By comparing the generation of UHMWPE particles with microfabricated surfaces in the water and in the bovine serum, the effect of biofluid lubrication on the particle size and shape will be elucidated. The results can assist the design of the surface textures on the articulating surface of the total joint implants.

2. Materials and methods A linear reciprocating wear tests was carried out in this study. Articulating materials are UHMWPE and textured surface of silicon. Original and highly cross-linked UHMWPE were tested under water and bovine serum lubricating conditions. The generated wear particles were isolated and characterized. The details of materials, wear process, particle collection and particle characterization procedures are described below. 2.1. Materials Raw GUR1050 UHMWPE and highly cross-linked GUR 1050 UHMWPE materials obtained from United Orthopaedic Corporation, Taiwan were used in this study. UHMWPE cylinder pins were machined to 6.35 mm in diameter and 25.4 mm in length with diamond turning on both end surfaces without polishing. The mean roughness (Ra) of UHMWPE pins’ end surface is 0.82 ␮m. Purified water (deionized, filtered with a 0.1 ␮m pore size membrane, and double distilled) and bovine calf serum were two lubricants used in the experiments. All UHMWPE pins were presoaked

281

in water or in serum for at least 15 days so as to become completely saturation with the immersed lubricant in the experiment. 2.2. Microfabrication of surface textures Texturing of the silicon wafer surface was achieved by photolithography patterning and etching of the bulk substrate [18]. A flow chart of the fabrication process is shown in Fig. 1. The detailed procedures and equipments have been described in a previous publication [16]. Two-inch diameter polished type P silicon wafers with (1 0 0) orientation were purchased from Summit-Tech and used as the base material for the texturing process. Wet oxidation of the silicon wafers was carried out in a glass-tube oven at 1100 ◦ C for 135 min to form a silicon dioxide film with a thickness of 1 ␮m. A pattern of rectangles with different size and aspect ratios were formed on a chrome direct-writing photomask. The silicon dioxide surface was spin coated with a Shipley 1813 positive photoresist. The dark-featured photomask was then placed on the photoresist surface and exposed to a UV source in a mask aligner to decompose the surrounding polymer surface leaving a positive rectangular pattern. The decomposed photoresist polymer was then removed in a Shipley 351 developer. The resulting positive photoresist surface was then etched first with a BOE solution (diluted buffer HF solution) to etch away the SiO2 in a wet chemical bath. The photoresist was then removed by washing with acetone-alcohol. Subsequently, the silicon material was subjected to isotropic silicon etching (HNA etchant; liquid volume ratio HF:HNO3 :CH3 COOH = 8:75:17) in a wet chemical bath at room temperature. The SiO2 layer was removed after an isotropic undercutting etching process. The resulting surface features are an array of rectangular ridges with sharp edges. Finally, a layer of 5 nm Cr coating was evaporated onto the surface to increase the strength and wear resistance of the surface texture. The height of the surface textures were measured by a Mahr profilometer and the feature length and width were measured from SEM observations. 2.3. Particle generation by a wear process A linear reciprocating wear test was applied to generate UHMWPE wear particles. ASTM F732 was used as a guideline [19]. The setup of the system is shown schematically in Fig. 2. One UHMWPE pin and three control pins were weighed three times. Three control pins were soaked in the lubricant at the same liquid level as the wear-testing condition. The UHMWPE pin was then mounted on the tester. Linear reciprocating wear tests were run under a nominal contact pressure of 3 MPa, a stroke length of 19 mm, a frequency of 1.5 Hz, and an average sliding speed of 57.2 mm/s for 4 h. After wear testing, the four pins were weighed and the wear loss of the testing pin was obtained after adjusting the weight change from the control pins.

282

H.-W. Fang et al. / Materials Chemistry and Physics 95 (2006) 280–288

Fig. 1. Schematic of the microfabrication process of surface textures on the silicon wafer.

2.4. Collection of wear particles in water After the wear tests, the UHMWPE particles were collected by repeated rinsing (purified water, adjusted to a pH 5.5 with hydrochloric acid) of the sample, sample holder, and parts that came into contact with particles, into a sterilized beaker. The procedure has been optimized to achieve a 90% particle recovery rate [16]. Purified isopropyl alcohol (0.1 ␮m filtered) was added and the particles were dispersed in an ultrasonic water bath for 3 min. 2.5. Isolation of wear particles in serum UHMWPE particles were collected by repeated water rinsing of the sample, sample holder, and parts that came into contact with particles, into a sterilized beaker [20]. In order to digest the proteins in serum, 5 mL collected serum solution was added with 20 mL NaOH solution (5N) in a 65 ◦ C water bath for 24 h. Three milliliters digested solutions were added with 2 mL glucose gradient solutions (5%, 10%, 20%, and 50%). The solution was then centrifuged at 4000 rpm and 4 ◦ C for 2 h [21]. The upper solid layer was collected and rinsed with 30 mL purified water. The solution was then agitated at 65 ◦ C for 1 h and followed by ultrasonification for 10 min to disperse the polyethylene particles. The above solu-

tion was then added 0.90 g/cm3 and 0.96 g/cm3 forgot thm3 isopropyl alcohol solutions with an amount of 2 mL, respectively, and followed by the centrifugation at 4000 rpm and 25 ◦ C for 1 h. The particles existing in the interface between two different isopropyl alcohol solutions were collected. 2.6. Characterization of particles Particles that were dispersed well in a known volume of the solution were collected on a 0.1 ␮m pore size membrane through a vacuum filtration process. The particles collected on the filter paper were examined by using a scanning electron microscope. Micrographs of the particles were then analyzed by using an image analyzer software (Scion Image, a PC ver-

Fig. 2. Experimental setup of linear reciprocating wear testing.

H.-W. Fang et al. / Materials Chemistry and Physics 95 (2006) 280–288

sion of NIH Image) to measure their dimensions. Measurements were made for at least 300 particles in each case. The chemical composition of collected particles was confirmed to be polyethylene by FTIR measurements (Nicolet) [16].

3. Results

283

Table 1 Surface-feature dimensions of microfabricated surface textures

Feature length (␮m) Feature width (␮m) Feature height (␮m) Distance between features (in sliding direction) (␮m)

Surface A

Surface B

Surface C

56.0 4.6 2.1 100

15.6 4.8 2.0 100

5.2 4.8 2.1 100

3.1. Microfabricated surface textures Surface textures with various dimensions of cutting edges and distances between features have been microfabricated. Fig. 3 shows the scanning electron microscopy (SEM) images of three surface textures used in this study. Table 1 lists the critical dimensions of the cutting-edges on the surfaces.

3.2. Effects of lubricants SEM images of wear particles of raw UHMWPE generated with microfabricated surface textures (surfaces A, B, and C shown in Fig. 3) under water and serum lubricating

Fig. 3. SEM images of microfabricated surface textures.

284

H.-W. Fang et al. / Materials Chemistry and Physics 95 (2006) 280–288

Fig. 4. Comparison of wear particles of raw UHMWPE generated in water and in serum with (a) surface A, (b) surface B, and (c) surface C.

conditions are shown in Fig. 4. Particles generated in serum were found to be smaller and more fiber-like than the particles generated in water. Elliptic major length and minor length measured from the image analysis represent the particle length and width in this study. The mean particle length and aspect ratio of wear particles generated in water and in serum are listed in Table 2. 3.3. Raw versus highly cross-linked UHMWPE The morphology of wear particles of raw and highly crosslinked UHMWPE generated with surface texture A in water are compared in Fig. 5. The morphology of wear particles of raw and highly cross-linked UHMWPE generated with surfaces A and B in serum are also compared in Fig. 6. Obviously, highly cross-linked UHMWPE wear particles have smaller sizes as well as smaller aspect ratios.

4. Discussion 4.1. Effects of lubricants on particle morphology It is observed that the mean particle length of raw UHMWPE wear particles generated in serum is much smaller than the particles generated in water. Fig. 7 compares the distribution plots of the length and aspect ratio of the wear particles generated with surface A in serum and in water. The mean particle length of the particles generated from surface A in water and in serum are 29.1 and 9.8 ␮m, respectively. 70% of particle length decrease occurs while the lubricant was switched from water to serum. It was also observed that by using surfaces B and C, the length of the particles generated in serum are about 10–20% of the particle length while using water as lubricants. These results all indicate that the particle size reduces when we changed

H.-W. Fang et al. / Materials Chemistry and Physics 95 (2006) 280–288

285

Fig. 5. Wear particles of raw and highly cross-linked UHMWPE generated in water (surface A).

Fig. 6. Wear particles of raw and highly cross-linked UHMWPE generated in serum (surfaces A and B).

the lubricants from water to serum. On the other hand, the particles generated in serum were found to be more elongated with larger particle aspect ratios as shown in Table 2. It was seen that the same trend was also observed on the highly cross-linked UHMWPE. For example, the mean particle lengths of highly cross-linked UHMWPE wear particles generated with surface B in water and in serum are

17.4 ± 5.3 and 1.0 ± 0.4 ␮m, respectively. The highly crosslinked UHMWPE particles generated in serum has a higher aspect ratio of 2.5 ± 1.3 than the aspect ratio of 2.1 ± 0.8 for the particles generated in water. In general, both lengths and widths of the particles are decreased and the morphology of the particles is more fiber-like under the serum lubricating condition.

Table 2 Particle length and aspect ratio of raw UHMWPE particles generated in water and in serum Surface A

Particle length (␮m) Aspect ratio

Surface B

Surface C

In water

In serum

In water

In serum

In water

In serum

29.1 ± 9.1 3.5 ± 1.5

9.8 ± 3.1 4.6 ± 2.2

18.1 ± 4.8 1.9 ± 0.7

3.4 ± 1.2 2.3 ± 1.1

7.4 ± 2.9 2.2 ± 0.5

0.8 ± 0.4 2.3 ± 1.0

286

H.-W. Fang et al. / Materials Chemistry and Physics 95 (2006) 280–288

tilting or twisting of the micro-cutting process and ends up with a smaller particle length. In addition, the high viscosity of the serum may make the possibility of generated particles adhered on the surface and leads to further re-processing or multi-micro-cutting of the generated wear particles. This may be one of the reasons causing the reduction of the particle length in serum. 4.3. Effects of degree of cross-linking on particle morphology Under both water and serum lubricating conditions, the wear particles of highly cross-linked UHMWPE are smaller and more round-like than the raw UHMWPE wear particles. It is known that the molecular chain of UHMWPE can be visualized as a tangled string of wires of C–C bonds (Fig. 9). The chain is not static and its arrangement is a function of temperature, applied stress, etc. For example, when a shear force is applied to the material surface, the molecules tend to align in the direction of shear stress [22], and the surface material strength may be strongly dependent on the direction

Fig. 7. Distribution plots of the length and aspect ratio of the wear particles generated with surface A in serum and in water.

4.2. Mechanism of UHMWPE particle generation with surface textures We should look into the particle-generation process in order to further understand the effects of biological lubricants on the wear particle morphology. The uniformed surface textures prepared by the microfabrication techniques has enabled the possibility to discuss the particle-generation mechanism from a single controlled asperity on the surface. A previous study has indicated that the micro-cutting mechanism of the wear particle generation with microfabricated surfaces consists of the following steps [16]: (a) penetration of the surface feature; (b) lateral displacement of the feature; (c) strain hardening of the polymer; (d) embrittlement of the surface layer by molecular orientation; (e) detachment of the particle at the tip edge–material interface. Fig. 8 is a schematic of the particle generation process with surface features. By using serum as a lubricant, the friction force existing in the interface of UHMWPE material and any single surface feature is reduced (see Fig. 8). The relative motion of the cutting-edge on UHMWPE materials becomes more “slippery”. This “slippery” phenomenon has two effects on the micro-cutting process of the particle generation. With a low frictional resistance, it maintains a higher net force pushing the cutting-edge away from the material surface. The time needed to complete a single particle generation is shortened. It results in the decrease of the cutting path in the lateral direction. Therefore, a smaller particle width is expected. Secondly, the slippery condition causes the minor

Fig. 8. Schematic of the particle generation mechansim with surface features.

H.-W. Fang et al. / Materials Chemistry and Physics 95 (2006) 280–288

287

modification of the synovial fluid. Thus further design of the surface pattern of the joint surface to control the size and shape of the wear particles may become possible. It is necessary to further obtain the correlations between the surface dimensions and particle morphology in order to establish the design guidelines of the articulating surface of joint implant to avoid particles-induced osteolysis.

5. Conclusions Fig. 9. Schematic diagram of the chain orientation of UHMWPE.

of shear motions. The highly cross-linked UHMWPE has a stronger molecular bonding. Higher degree of cross-linking may reduce the degree of molecular orientation change under applied stresses. Under a linear reciprocating wear test, the x-PE may behave more brittle and the material tends to break easily. Therefore, highly cross-linked UHMWPE has a better wear resistance and a more isotropic material properties. Under a micro-cutting process, smaller and more round wear particles were observed on the highly cross-linked UHMWPE regardless using water or serum as lubricants.

To sum up, by using the microfabricated surface textures, a narrow distribution of the particles has been achieved. It demonstrates the ability to control the particle size and shape by the dimensions of the surface textures. It is observed that the length of the particle generated in serum is smaller than the particle generated in water. The particle generated in serum has a larger aspect ratio. Under the lubrication of bovine serum, both length and width of the wear particle shrinks. We have preliminarily observed the effect of biofluid lubrication effect on UHMWPE particle morphology with microfabricated surface textures.

Acknowledgements 4.4. Tribochemistry issues The role of biological lubricants on controlling of the wear particle morphology should be further illustrated in order to facilitate the surface texturing techniques on the articulating surface of artificial joints in vivo. The major components of the synovial fluid of the joint (biological lubricant) include albumin, globulin, lipids and hyaluronic acid. Under the articulating of the joints, the synovial fluid is squeezed and forms a boundary molecular layer responsible for the lubricating of the joints. The adsorption strength of the molecules and the tribochemical reaction of the molecules are the factors affecting the wear and wear particle formation of the materials. Liao et al. has studied the thermal effects on the serum protein contents and [23] Heuberger et al. has discussed the modification of the albumin conformation due to the tribological thermal effect [24]. The adsorption of the related proteins on articulating surface has been investigated and measured [24,25]. Wong et al. indicated that the micro-cutting phenomenon of particle generation is an abrasion process related to the lowcycle fatigue [26]. However, the biological lubricant plays a role transferring the friction work to the materials during the articulation. Different physical and chemical properties of lubricant may influence the amount of the work transferred to the materials and contribute to the low cycle fatigue. Thus the variance of particle size and shape can be expected due to different biological lubricant used. However, by introducing the surface textures to the articulating surface of the joints, uniform normal and shear stress distribution are created on the material surface, we need to further investigate the thermal and mechanical effects on the

The authors are pleased to acknowledge the financial support of the National Science Council (NSC 92-2218-E-027013-, and NSC 93-2213-E-027-010-), Taiwan. Special thanks to Precision Instrument Development Center, Hsin-Chu, Taiwan for their assistance on microfabrication processes.

References [1] W.H. Harris, The problem is osteolysis, Clin. Orthop. Res. 311 (1995) 46–53. [2] H.C. Amstutz, P. Campbell, N. Kossovsky, I.C. Clarke, Mechanism and clinical significance of wear debris-induced osteolysis, Clin. Orthop. Res. 276 (1992) 7–18. [3] W.J. Maloney, R.L. Smith, Periprosthetic osteolysis in total hip arthroplasty: the role of particulate wear debris, J. Bone Joint Surg. 77A (1995) 1448–1461. [4] T.R. Green, J. Fisher, M. Stone, B.M. Wroblewski, E. Ingham, Polyethylene particles of a ‘critical size’ are necessary for the induction of cytokines by macrophages in vitro, Biomaterials 19 (1998) 2297–2302. [5] A.S. Shanbhag, C.T. Hasselman, H.E. Rubash, Technique for generating submicron ultra high molecular weight polyethylene particles, J. Orthop. Res. 14 (1996) 1000–1004. [6] M. Allen, F. Brett, P. Millett, N. Rushton, The effects of particulate polyethylene at a weight bearing bone-implant interface, J. Bone Joint Surg. 78B (1996) 32–37. [7] S.B. Goodman, P. Huie, Y. Song, D. Schurman, W. Maloney, S. Woolson, R. Sibley, Cellular profile and cytokine production at prosthetic interfaces, J. Bone Joint Surg. 80B (1998) 531–539. [8] D.W. Howie, B. Vernon-Roberts, R. Oakeshott, B. Manthey, A rat model of resorption of bone at the cement-bone interface in the presence of polyethylene wear particles, J. Bone Joint Surg. 70A (1998) 257–263.

288

H.-W. Fang et al. / Materials Chemistry and Physics 95 (2006) 280–288

[9] I. Catelas, O.L. Huk, A. Petit, D.J. Zukor, R. Marchand, L. Yahia, Flow cytometric analysis of macrophage response to ceramic and polyethylene particles: effects of size, concentration, and composition, J. Biomed. Mater. Res. 41 (1998) 600–607. [10] S.M. Horowitz, J.B. Gonzales, Effect of polyethylene on macrophages, J. Orthop. Res. 15 (1997) 50–56. [11] I. Voronov, J.P. Santerre, A. Hinek, J.W. Callahan, J. Sandhu, E.L. Boynton, Macrophage phagocytosis of polyethylene particulate in vitro, J. Biomed. Mater. Res. 39 (1998) 40–51. [12] M.E. Martinez, S. Medina, M.T. del Campo, J.A. Garcia, A. Rodrigo, L. Munuera, Effect of polyethylene particles on human osteoblastic cell growth, Biomaterials 19 (1998) 183–187. [13] A.S. Shanbhag, J.J. Jacobs, J. Black, J.O. Galante, T.T. Glant, Macrophage/particle interactions: effect of size, composition, and surface area, J. Biomed. Mater. Res. 28 (1994) 81–90. [14] H.-W. Fang, S.M. Hsu, J.V. Sengers, Surface texture design to generate specific sizes and shapes of UHMWPE wear particles, Materialwissenschaft und Werkstofftech (Mater. Sci. Eng. Technol.) 34 (2003) 976–988. [15] H.-W. Fang, S.M. Hsu, J.V. Sengers, Generation of narrowly distributed ultra-high molecular weight polyethylene particles by surface texturing techniques, J. Biomed. Mater. Res. B Appl. Biomater. 67B (2003) 741–749. [16] H.-W. Fang, S.M. Hsu, J.V. Sengers, Ultra-high Molecular Weight Polyethylene Wear Particle Effects on Bioactivity, vol. 1002, National Institute of Standards and Technology Special Publication, 2003. [17] S.-Y. Yang, W. Ren, Y. Park, A. Sieving, S. Hsu, S. Nasser, P.H. Wooley, Diverse cellular and apoptotic responses to variant shapes of UHMWPE particles in a murine model of inflammation, Biomaterials 23 (2002) 3535–3543.

[18] W. Ren, S.-Y. Yang, H.-W. Fang, S.M. Hsu, P.H. Wooley, Receptor activator of NF-␬B (RANK) and RANK ligand (RANKL) expression in response to varying morphology of UHMWPE particles in a murine model of inflammation, Biomaterials 24 (2003) 4819– 4826. [19] ASTM F732, American Standards of Testing Materials, 1996. [20] P. Campbell, S. Ma, B. Yeom, H. McKellop, T.P. Schmalzried, H.C. Amstutz, Isolation of predominantly submicron-sized UHMWPE wear particles from periprosthetic tissues, J. Biomed. Mater. Res. 29 (1995) 127–131. [21] S. Affatato, D. Emiliani, G. Bersaglia, F. Traina, A. Toni, An easy technique to digest and isolate wear particles from a hip joint simulator, J. Artif. Organs. 27 (2004) 424–432. [22] S. Sambasivan, D.A. Fischer, M.C. Shen, S.M. Hsu, Molecular orientation of ultrahigh molecular weight polyethylene induced by various sliding motions, J. Biomed. Mater. Res. B Appl. Biomater. 70B (2004) 278–285. [23] Y.S. Liao, D. McNulty, M. Hanes, Wear rate and surface morphology of UHMWPE cups are affected by the serum lubricant concentration in a hip simulation test, Wear 255 (2003) 1051– 1056. [24] M.P. Heuberger, M.R. Widmer, E. Zobeley, R. Glockshuber, N.D. Spencer, Protein-mediated boundary in arthroplasty, Biomaterials 26 (2005) 1165–1173. [25] T.A. Blanchet, S.L. Peterson, K.D. Rosenberg, Serum lubricant absorption by UHMWPE orthopaedic bearing implants, ASME J. Tribol. 124 (2002), 12/29/20041-4. [26] B.K.P. Wong, S.K. Sinha, J.P.Y. Tan, K.Y. Zeng, Nano-wear mechanism for ultra-high molecular weight polyethylene (UHMWPE) sliding against a model hard asperity, Tribol. Lett. 17 (2004) 613– 622.