Mechanical Properties of Processed Ultra High Molecular Weight Polyethylene
Bioengineering Laboratory 210
Final Project
April 29, 1998
Group R6
Amie Borgstrom
Louis Kolman
Robert Ledger
Maryam Malik
Mechanical Properties of Processed Ultra High Molecular Weight Polyethylene
Abstract
Changes in density, percent crystallinity, and elastic modulus caused by fabrication processes were measured in ultra high molecular weight polyethylene to determine if surface oxidation occurred during fabrication. The samples of polyethylene used for testing included a non-sterilized compression molded block, inner and surface regions of a gamma irradiated knee implant, and a resin in powder form. Densities of the irradiated samples fell within the accepted ASTM density standard of 0.930-0.944 g/cm3. Compression molding caused a decrease in percent crystallinity between the resin and block form (86.51% to 52.40%). Gamma irradiation and further machining caused the crystallinity of the material to increase to 59.58%. All calculated elastic moduli agreed within 7.5% of the 703 MPa elastic modulus experimentally determined by Deng et al. No statistical difference was found between the densities, percent crystallinities, or elastic moduli of the interior and surface of the irradiated material. No discoloration due to oxidation was observed under the light microscope (20x).
Results implied insignificant oxidation in these samples. Furthermore, no conclusion could be made solely on the effects of gamma sterilization, as there was no control sample against which to compare the irradiated sample.
Background
Ultra-high molecular weight polyethylene (UHMWPE) is a linear polyethylene that has an extremely high molecular weight (approximately 4 x 10^6 g/mol). UHMWPE posses an array of properties including high abrasion resistance, low friction, high impact strength, low density, excellent toughness, ease of fabrication, biocompatibility and biostability. It is these characteristics that have made UHMWPE particularly attractive for use in designing bearing surfaces in arthroplasties (Callister, 499). UHMWPE is currently the leading material used for the manufacture of the liner of the acetabular cup in total hip arthroplasties (THAs) and the tibial insert and patellar components in total knee arthroplasties (TKAs). Clinical performance of these components is considered to be fairly successful except for the concern of wear and creep. The primary concern is not that the components will wear out, but that the wear debris will cause a series of undesirable effects. Even after the elimination of most of these particles by the body’s host response, large numbers of them still remain in the body tissues (Red Journal). According to an article by W.J. Maloney, these particles are phagocytosed, resulting in many reactions such as the formation of grannulamatous lesions, osteolysis, bone resorption, and loosening of THAs and TKAs (Maloney, 1449-1450). These possible phenomena have caused the problem of UHMWPE wear to be one of the most challenging problems in contemporary orthopedics. UHMWPE is a viscoelastic material and has a lower elastic modulus compared to those of metal implants and cortical bones, making it susceptible to creep deformation. Creep of a polyethylene implant could lead to thinning of the material at the region where a majority of the load is transferred and result in accelerated wear of the articular surface (Callister, 481).
The most important endogenous manufacturing factors influencing the properties of UHMWPE are the nature or quality of the starting powder, the method of manufacture of the component, the component sterilization conditions, and the storage environment. The physical properties of the starting resin are extremely important as differences in the bulk resin material properties can affect the performance of the final product. For instance the presence of calcium stearate has been shown to cause morphological defects according to Zimmer. There are several processes of choice for fabricating these components including ram extrusion, direct compression molding, and heat pressing. Direct compression molding, the one-step method used for the samples tested in this laboratory experiment, involves placing the powder in a heated mold and applying pressure to form the component (Callister, 492). Although compression molded parts require only one step, they have characteristic defects. For example, large variations in crystalline morphologies are sometimes present in these parts (Zimmer).
After the material has been fabricated, it must be sterilized and shaped for implantation into the body. Gamma radiation is a common sterilization method used in materials for total joint replacement. This method has many advantages as it is a fully penetrating sterilant and has a low cost. It has been found that when polyethylene components are exposed to energy in the form of gamma irradiation, free radicals are formed due to the breaking of carbon-carbon bonds or carbon-hydrogen bonds. This phenomenon leads to polymer chain scission, free radicals do not recombine and instead combine with oxygen, and chain cross-linking, free radicals recombine to form a three dimensional structure (Rimnac, 1052-1056). Some researchers believe that this formation of free radicals leads to the subsequent surface and subsurface oxidative degradation of the virgin component. Surface oxidative degradation results from the interaction of oxygen from the atmosphere with the long-lived free radicals. Subsurface oxidative degradation is a consequence of the reaction of these radicals and the oxygen that has diffused into and dissolved in the polymer. These degradation mechanisms are transient diffusion processes and thus are component shelf-life dependent. There is much agreement on the aspects of the formation of free radicals described above. However there is an ongoing controversy regarding whether the chain scission or cross-linking is the dominant result of gamma radiation. There is also much controversy surrounding the occurrence of oxidation (Sun, 22-26).
Companies in support of gamma radiation such as Zimmer, Inc. claim that by gamma sterilizing polyethylene in a reduced oxygen environment (nitrogen processed and packaged), free radicals are able to recombine rather than react with oxygen, leading to increased cross-linking. They state that this increased cross-linking tends to lead to a higher molecular weight because the chain is longer, thus there is a greater probability that a carbon atom will produce a secondary branch to begin another chain. Researchers in favor of gamma- sterilization state that increased cross-linking tends to lead to increased wear resistance and will enhance other mechanical properties (Zimmer).
Researchers not in favor of gamma sterilization claim that sterilizing polyethylene in an inert environment will help to reduce the initial oxidation that occurs, but cannot eliminate oxidation due to the presence of dissolved oxygen in polyethylene. Thus the oxidation of UHMWPE continues for long periods of time following gamma radiation. UHMWPE initially consists of extremely long molecular chains, a characteristic that allows for excellent abrasion resistance. They claim that in sterilization by gamma radiation, chain scission is dominant. (splitting of hydrogen-carbon bond from chain) (Wright Med. Tech). Oxygen diffuses into the material and reacts with free radicals (single bond with carbon) to cause oxidation, leading to shorter molecular chains, a lower molecular weight, embrittlement, and decreased wear resistance. The altered properties of the polyethylene can adversely affect its performance in joint replacement (Ries, 757).
These differing conclusions can be tested by looking at the percent crystallinities of the polyethylene samples, found using the Differential Scanning Calorimeter. Measurements taken using the DSC are based upon the amount of heat absorbed by the material during a thermodynamic reaction caused by a phase transition. Values for the heat of fusion of the samples are given directly by the Perkin Elmer Data Analysis program (using an estimation of the area under the melting transition peak). From the heat of fusion values the percent crystallinity can be calculated using the formula:
% Crystallinity = (H) where 290 J/g is the heat of fusion of purely (1)
(290J/g)crystalline UHMWPE (BE 210 Lab Manual)
The amount of energy required to melt a crystalline molecule is greater than the amount of energy needed to melt an amorphous molecule, as the bonding in the crystalline material is more organized and has more bonds. Thus, heat of fusion is directly proportional to crystallinity. This can also be explained by looking at the relation between density and heat of fusion. A more dense material will have more bonds that exist to absorb heat during melting, giving a higher heat of fusion. (Only the crystalline portion of the sample has a well-defined melting point and heat of fusion while the values for the amorphous portion may vary.) (BE 210 Lab Manual). Ideally density is directly proportional to crystallinity, but this relation becomes distorted in a polymer with chain scission and cross-linking. Density will be tested using a density gradient column (Zimmer).
The above differing conclusions can also be tested using a three-point bending machine as the Young’s Modulus of polyethylene is sensitive to oxidation. In order to calculate the elastic modulus (Pa) of each sample the following formula was used:
E= - 3 PL2 where P = maximum force (N) (2)
48 bh3y L = length (meters)
b = base (meters)
h = height (meters)
y = maximum displacement (meters)
(BE 209 Lab Manual)
The bonds in the crystalline region are more likely to fracture under high load instead of elongating. Testing has also shown extensive reduction in toughness and ductility, primarily in the surface layer due to oxidation, thus increasing the inclination to failure. According to William Callister an increase in cross-linking will inhibit relative chain motion and therefore strengthen the polymer. Secondary intermolecular bonds (van der Waals) are effective in this inhibiting of chain motion, thus, the mechanical properties of polymers are highly dependent on their magnitude. Furthermore, the degree of crystallinity will have a significant influence on the mechanical properties of the polymer since it affects the extent of these secondary bonds. A dominant increase in crystallinity over density (implying chain scission, which leads to oxidation), typically leads to an increase in the modulus of elasticity. This intensifies contact stresses. Undispersed contact stress can contribute to UHMWPE wear (Callister, 467).
Researchers at Zimmer claim that gamma sterilization affects both density and crystallinity. They state that a significantly higher percent increase in density relative to crystallinity implies dominant cross-linking, while a higher percent increase in crystallinity implies dominant chain scission. The information given by Zimmer was obviously taken with skepticism, as they are indeed trying to make their product appear superior. However, the above statements pertaining to crystallinity and density were assumed not to have been falsified in any way (Zimmer).
In this experiment a UHMWPE sample from each of the three different processing stages (in preparation for implantation in the human body) were tested for changes in density, crystallinity, and elastic modulus. The knee sample was also examined for oxidation under a light microscope. The samples (resin, compression-molded block, sterilized implant – inner section, sterilized implant - surface section) were produced by Zimmer, Inc. Zimmer’s approach to fabrication is to start with high quality polyethylene raw material resin, free of calcium stearate. The knee implant is then compression molded, in a slab-molded form, followed by machining to final component geometry. Zimmer states that this form of processing is used in order to insure optimal polymer fusion and to closely maintain the original mechanical properties (relative to changes caused by other forms of processing). Polymer fusion is important, as shrink voids and other morphological defects have been implicated as detrimental factors to wear rates and in vitro mechanical properties. Zimmer then molds it into the implant shape, sterilizes the product via gamma-irradiation, and packages its UHMWPE in a nitrogen environment (Zimmer).
Apparatus and Materials
It should be noted that all the samples used in this analysis are the same material (ultra high molecular weight polyethylene) taken from different stages of processing. All samples required cutting. For the differential scanning calorimeter and the optical light microscope, a scalpel was used to obtain small pieces of polyethylene. Samples used in the three point bending machine required precise measurements of working length (5/8), thickness (1/8), and width (1/4). These samples were sent to a machine shop for cutting. The samples were provided by Zimmer, Inc., a maker of prosthesis parts.
[1] ultra high molecular weight polyethylene resin: the sample provided in the powder form was used in the Differential Scanning Calorimeter to obtain a percent crystallinity.
[2] ultra high molecular weight polyethylene compressed molded block: (non-sterilized) sample was cut and used for analysis in the optical light microscope, differential scanning calorimeter , and three point bending machine.
[3] ultra high molecular weight polyethylene -irradiated sterilized knee implant: sample was cut to obtain an inner region and outer surface and used for analysis in the optical light microscope, differential scanning calorimeter , and three point bending machine.
[4] Perkin-Elmer AD-4 Autobalance: used to obtain a precise mass of the polyethylene samples prior to initial experimental testing.
[5] Aluminum pans and lids: used to contain the polyethylene samples while being analyzed in the differential scanning calorimeter.
[6] Instron 3116 Three Point Bending Machine: used to determine the elastic moduli of the polyethylene samples.
[7] Perkin-Elmer Universal Analysis Program: used to analyze the data collected from the differential scanning calorimeter. Specifically, the program is used to determine the onset of melting, the melting point, and the percent crystallinity of the polyethylene sample being analyzed.
[8] Perkin Elmer Differential Scanning Calorimeter: used to measure the heat of fusion of the polyethylene samples over a specific range of temperatures. The heat of fusion was determined to compare the percent crystallinity of the samples.
[9] Olympus SZ-60 Optical Light Microscope: physical appearance of the -irradiated sterilized UHMWPE knee implant was observed for discoloration in order to test for oxidation occurring at the surface.
[10] Density Gradient Column: prepared using mixture of water and isopropanol and used to determine the specific density of each of the samples.
Procedure
Differential Scanning Calorimeter (DSC)
The heat of fusion of each sample was measured using the DSC. Small samples (n=3) of UHMWPE weighing between 5mg and 10mg were cut from each of the materials. The samples were massed and placed in aluminum pans with covers. The small sample were sealed in the aluminum containers and placed in the DSC. The samples were first held at 25C for 30 seconds and then heated at a constant rate of 10C/min until they reached 170C. Heating was carried out in a nitrogen environment. Results were recorded with the Perkin-Elmer Analysis Program. A plot was constructed of heat flow (mW) versus temperature (C). (Appendix Figure 3) The heat of fusion was calculated by the program as the area under the heat flow curve. The area under the curve was taken from the limits of 30C to 152C for each curve. Tests were repeated in similar fashion for the resin, the inner region of the irradiated knee, and the surface of the irradiated knee (n=3). Percent crystallinity was then calculated from the heat of fusion values.
3 Point Bending
Elastic modulus was measured using a 3 point bending test. Samples (n=1) were cut from the inner region of the irradiated knee, the surface region of the irradiated knee, and the block. The samples were cut into rectangular blocks to dimensions of approximately an 1/8” inch in height, 1/4” in width, and 5/8” length. The exact dimensions of each sample were measured with a caliper prior to testing. Loads of 10 lb. (44.48 N) were applied to each sample at a rate or 0.5 lb/sec. Force and displacement were recorded in plotted against each other. The elastic modulus was calculated using Equation 2. Three trials were executed on each sample.(Appendix Figure 2)
Optical Light Microscope
The irradiated knee sample was observed under the light microscope for discoloration due to oxidation. Prior to observation, the cross-sectional surface of the sample was polished using a silicon carbide paper in order to create a flat, clean surface. The surface of the sample was examined up to a depth of 8 mm. Oxidation is most prevalent at depths of 3-4mm (Zimmer).
Density Gradient Column
A density gradient column was constructed using a mixture of isopropanol and water (Figure 1). A sealed tube with a clamp connected two large flasks. Flask A contained a mixture of 240 mL water and 700 mL isopropanol. Flask B contained 940 mL of water only. A long sealed tube with a clamp was run from the very bottom of the column to be filled and connected to the flask containing the mixture. The clamp between the column and the mixture was slowly opened to allow a low flow of liquid. As the liquid reached half way down the gradient column, the clamp between the two flasks was slowly opened to allow some water to mix with the alcohol/water mixture. The column was allowed to fill over the period of a week. Glass reference balls of known density were placed in the filled column. The density of the height of the balls was plotted versus their known density and placed with a linear fit. The equation of the line gave the relationship of height versus density.