Background:
The Effects of Gamma Radiation
on Ultra-High Molecular Weight
Polyethylene Used in
Implant Devices
BE210
Group W4
Kevin Justice
Joanna Law
Seungtaek Lee
Emily Rothman
Wednesday, April 30, 1997
Section Page
1. Abstract 3
2. Background 4
• properties of ultra-high
molecular weight polyethylene 4
• prosthetic design 7
• effects of gamma radiation
sterilization on UHMWPE 12
• differences between gamma irradiation in air
and inert environments 18
7. Methods and Materials 23
8. Results 28
• Differential Scanning Calorimeter 28
• Mechanical Testing 35
11. Discussion 38
12. References 46
13. Appendix 49
Two molded samples of ultra high molecular weight polyethylene used as bearing materials in total knee replacements were obtained from Zimmer (Warsaw, Indiana). Each sample had been subjected to a different technique of gamma irradiation sterilization: one packaged and irradiated in an inert atmosphere, the other in air. It was sought to determine both qualitatively and quantitatively the effects of the two procedures on the physical and mechanical properties of the UHMWPE. The samples were tested with the aid of the Perkin-Elmer DSC-7 to determine the heat of fusion and the percent crystallinity of the two samples. Additionally, an Instron testing machine Model 1331 was used to conduct mechanical testing, which included three point bending and a pseudo-Brinell hardness test. The value of heat of fusion of the air-irradiated sample was determined to be statistically constant at an average value of 117.31 J/g ( 3.93%. The inert-irradiated samples showed a trend for the heat of fusion to increase from an initial value of 104.47 J/g ( 3.0% to a value measured in the third week as 125.12 J/g ( 10.6%. It was further found that the modulus of elasticity as determined from the three point bending testing was not significantly different between the two sterilization techniques. Within experimental error, the mean Young’s Modulus was calculated as 1142.78 MPa ( 153.22 MPa. Further, the hardness testing showed that the air-irradiated sample was harder by an average 59.46% with a deviation 11.88% over a range of loading of 10 N through 190 N. Our findings, while limited by the brevity of the study, show fairly conclusively that the irradiation of UHMWPE in air leads to oxidative degradation, as evidenced by changes in heat of fusion, percent crystallinity and hardness (britttleness).
Properties of UHMWPE:
In total joint replacements, the current material of choice for use as a bearing surface is ultra-high molecular weight polyethylene. Joints made from this material can function for more than 15 years, but debris caused by wear makes longer term use impossible. Many other polymeric materials such as polytetrafluoroethylene, polyacetal, high-density polyethylene, polyesters, and carbon-reinforced ultra-high molecular weight polyethylene have been tested in joint replacements. However, none have proved to be as wear resistant as ultra-high molecular weight polyethylene. UHMWPE is often confused with high density polyethylene, but they differ in the properties of molecular weight, density, impact strength, and abrasive wear. UHMWPE has a density value which is closer to that of low-density polyethylene. However, not only does it have a high tensile strength and modulus, it also has a much higher molecular weight, higher impact strength and toughness, and better characteristics of abrasive wear than high-density polyethylene.
More than 90,000,000 pounds of ultra-high molecular weight polyethylene were produced in 1993, but the main industrial uses are ski-bottoms, cutting boards, and coal-chute liners. About 400,000 total joint replacements were performed the same year, and one-quarter pound of polymer was used for each joint. So, 100,000 pounds of polymer was implanted which is 0.1% of total annual production of UHMWPE. Thus, since medical uses of the polymer consume such a small portion of the total produced, manufacturers have paid little attention to the specific needs of the orthopedic community until recently.
Polyethylene is a long chain of monomer units of ethylene from which side groups branch off and become attached to the main chain. The molecular weight of UHMWPE is the molecular weight of ethylene multiplied by the number of ethylene groups that make up one molecule of polyethylene polymer. This molecular weight is determined by measuring the relative viscosity of the material (by ASTM standards) and using this value to calculate the intrinsic viscosity. Then, using the Mark-Houwink equation, the molecular weight is a function of the viscosity raised to the power of 1.7. Thus, because the relationship is exponential, small changes in measured viscosity result in large changes in molecular weight. Consequently, the reported molecular weight is anywhere within the range of 2 to 6 million.
Ultra-high molecular weight polyethylene has both an amorphous and a crystalline phase. The percentage of the material that is crystalline is limited because the long chain lengths hinder the ability of molecules to form crystalline arrays. The percentage of crystallinity can be measured using differential scanning calorimeter (DSC). This determination works by the gradual application of heat to the sample, and measurement of the amount of heat absorbed. Heat of fusion is determined by measuring the area under the curve. 100% crystalline polyethylene has a heat of fusion of 288 J/g, so (Hf sample/ 288) x 100 gives the percent crystallinity of the sample. Synthesized UHMWPE powders have crystallinity values ranging from 58% to 75%, depending on the type of resin.
Physical properties, such as how the polymer responds to applied stresses and chemical environments, are determined by the morphological characteristics. The crystalline regions of UHMWPE (and any polymer) are made up of unit cells which group together to form lamellae. Although there have been recent reports of a spherulitic structure in UHMWPE, earlier studies showed that there is no formation of lamellae into spherulites in UHMWPE, as there is in other polymers. The lamellae are composed of folded polyethylene chains, and the unit cells are reported to be orthorhombic. In addition to the crystalline and amorphous phases, there is also an intermediate phase, which is hexagonal.
Unfortunately, physical defects have been observed in many new and retrieved ultra-high molecular weight polyethylene devices. These defects are small, round particles (between 50 and 200 micrometers thick) of polyethylene that do not adhere to the surrounding particles. Sometimes these defects are abundant and distributed evenly throughout the sample, but other times the particles are concentrated in different areas. The size of these particles is consistent with the range of diameters of polyethylene powder used to make the devices. After these particles are melted in a test at 140-145 degrees Celsius, they coalesce with the surrounding matrix when cooled. However, it is not known why these non-cohesive particles exist in UHMWPE devices.
The recent goal of many investigators is to alter the properties of UHMWPE by altering the material’s morphological characteristics. Although this is difficult because of the long lengths of the polymer chains several researchers have shown that the crystallinity of polyethylene can be increased without causing degradation by applying high temperatures and pressures. These forms with higher crystallinity are thought to exhibit a higher strength and modulus, but the methods used to produce it are not yet feasible commercially because of cost and scale limitations. Another new process has recently been developed to produce a range of higher-crystallinity forms of UHMWPE. This process allows recrystallization of solid UHMWPE from 50% to 80% crystallinity without degradation of the original material. The modulus of elasticity increases as much as 375% , and tensile yield strengths increase as much as 30%. However, the disproportional increase in modulus compared with yield strength causes a disadvantage in total joint designs that have subsurface shear stresses.
Prosthetic Design
In today’s society, there is a growing trend towards the development and implementation of artificial implants. More specifically, joint implants are being used in a larger percent of the geriatric population, especially in those who suffer from osteoporosis. More commonly, the knee, hip, shoulder and ankle are joints that often require replacement; with the knee and hip being apparent across all age groups.
However, in the implant, a great deal of research and studies must be done before a device may be considered suitable in the human body. The materials chosen are of great concern as is the shape and size of the implant. For example, in choosing the type of metal to be used, researchers must remember the aggressive environment of the human body. This plays a large factor as many metals cannot be used because of corrosion due to the ionic particles in the blood. The result of the ions present in the blood may cause galvanic corrosion which will result in the degradation and eventually failure of an implant. In addition to the metal choice, there must be some polymer present to take on the wear of the joint to ease the pain and movement of the two metal parts connecting the joint. In the choice of this polymer, the high amount of usage and motion of the daily activities of the human body must be taken into consideration as the mechanical properties and wear life of this polymer now plays a great role. But in addition to these two aspects, designers must take into consideration the shape and size of the polymer piece, that is, whether or not it should be flat or have a contour similar to that of the use of cartilage in the natural joints. Furthermore, in addition to these design limitations, there must be considered the method of sterilization of the implant and the method of importing the implant into the human body. In the stages of having a functional implant in the body, for usually anywhere between five to ten years at least, there are many stages and considerations that first must be considered.
In this project, we investigated the use of an ultra-high molecular weight polyethylene implant used in tibial inserts used for knee implants. Prior to doing tests on these implants, however, we first have to understand what we must be looking for in the analysis of present components that have failed in patients and the means they have failed in the human body. In a study done by Wasielewski, et al, twenty inserts of ultra-high molecular weight polyethylene and thirty-five were carbon-reinforced polyethylene were retrieved and examined for evidence of wear. The method that was followed in examination of the implants involved topographic maps of the articular and metal-backed surfaces of each component, in addition to which, radiographs of the pre- and postarthroplasty or prerevision inserts were analyzed for the component positioning, sizing and extremity alignment. However, it is the study of the polyethylene wear in the knee arthroplasties that give us the most information in terms of the design of the implant and the factors associated with it.
Knee implants have been recommended as a means of effectively eliminating pain and retaining motion in patients with pain in their knee and to recreate normal knee stability and function. In order to do this, though, all ligament abnormalities must be corrected in order to balance at arthroplasty so as to interact cohesively with the prosthetic joint to provide the stability and allow the knee its natural range of motion, the goal of the implant. Incompatibilities between implant geometry, orientation, or limb alignment and ligament may cause high loading or excessive knee laxity within the functional range of motion. These mechanical incompatibilities have been the source of implant failure (Wasielewski, et al.) and have also been discussed in relation to insert wear. However, it is the wear that we are most concerned with in terms of the study of the knee implant, as this is the result of multiple factors that eventually lead to the implant failure. Factors affecting wear include the time of implantation in vivo, polyethylene thickness, polyethylene voids, and polyethylene material properties.
An underlying reason that these factors affect the wear is the design and shape of the implants. In a study done by Bartley, et al., three different types of polyethylene articulating implants were studied. The three types (Porous Coated Anatomic, Howmedica; Fibermesh, Zimmer, and Microloc, Johnson & Johnson) differed not only in the insert shapes, but also in the method of manufacturing the polyethylene, in the design of the articular surfaces, and also in methods of insertion (i.e. bone cement or not). Table 1 lists the differences in the three implants, showing a comprehensive range of properties in a small, yet significant piece, in the knee implant. This goes to show that there is so much that goes into the design specifications and the research of an implant; in the formation of a piece in the knee implant, consideration must be taken into the type of polymer used, the size, shape, manufacturing, and method of insertion all vary and have the freedom to be different. Yet, there is apparent failure across all three designs. And this is not to say, though, that these always fail or that there is only one design out there that is perfect. Rather, it is dependent upon a great deal of factors that are too many to list. It can fail because the patient is too old or too young, too active or overweight, even if the patient has a different diet and therefore, a different pH content from another patient with the exact same
|Type of Implant |Polyethylene Insert Shape |Method of PE Manufacturing |Design of Articular surfaces |
|(Manufacturer) | | | |
|PCA (Howmedica) |non modular |heat pressing technology |replicate normal articular surfaces |
|Fibermesh (Zimmer) |non modular |machined |femoral component: |
| | | |round mediolaterally & anteroposteriorly |
| | | |polyethylene: flat medio-laterally & |
| | | |anteroposteriorly |
|Microloc |modular |machined |femoral component: flat mediolaterally & |
|(Johnson & Johnson) | | |rounded anteroposteriorly |
| | | |tibial component: flat medio-laterally & |
| | | |anteroposteriorly |
implant. However, some of these differences, though, may be controlled or at least observed for failure and documented to prevent any harm from occurring to the patient.
In all three types, though, polyethylene thickness and polyethylene modulus were associated with failure but these were only secondary to the importance and initiation and progression of the wear process (Bartley, et al) . It is only after the device has failed are we able to learn the most and make the most important changes to the design of the implant. From the knee implants, we learned that subluxation (the movement of the implant within the body after implantation so that it is no longer aligned with the natural bones and therefore, not allowing natural and free movement of the leg) is the main source of concentrated forces on the polyethylene where it is thinnest, thus causing wear and resulting in failure. This may be a direct result of the design of the polyethylene implant, or rather, it may also be a result of the material properties of polyethylene. Taking it to another level, it may have nothing to do altogether with the polyethylene and rather, with the design of the prosthetic as a whole, resulting in the subluxation and the result of the poor design results in the wearing of the polyethylene articulating surface. For example, in the brand in which there was perfect conformity between the femoral and tibial articular surfaces (PCA), the slopes of the surfaces became inherently unstable and therefore allowed early implant subluxation. As a result of the subluxation, the forces were concentrated at periphery of the tibial component where the polyethylene implant is thinnest (Bartley et al.). However, the other two did not fare so well either. These implants were designed with the flat tibial articular surfaces mediolaterally and anteroposteriorly. This resulted in the high contact stresses within the polyethylene due to the non-conformed geometry.
In this way, only at the expense of others, do we learn how to improve and truly design an implant that is accepted by the human body mechanically, chemically, and physically. It is through the careful observations and well-documented accounts of doctors and researchers to understand how the human body works, because in trying to make someone’s life better, there are many more considerations and factors that must be observed.
The Effects of Gamma Radiation Sterilization on UHMWPE
All biomaterials including UHMWPE and bio-medical devices must be sterilized before they can be implanted in the body. At present, the most common method of sterilization for heat sensitive products is through gamma radiation. The photon emitted from the radioactive substance will penetrate into the UHMWPE and kill any microorganisms that might react in vivo. Unfortunately there are negative side effects which arise from this sterilization technique. As previously mentioned, UHMWPE has great fracture, fatigue, and wear properties. However, when the material is sterilized with gamma radiation, the chemical and mechanical properties of these UHMWPE are altered. These changes continue both in shelf aging and in vivo use. These alterations can lead to degradation, loosening, pitting, delamination, formation of debris, and other forms of failure. Gamma radiation sterilization seems to be a stepping stone to a series of structural changes which ultimately results in a failure of UHMWPE implant.
As a result of the gamma radiation, free radicals are produced which will ultimately lead to the chemical transformation of the polyethylene. The source of these high-energy photons are usually Cobalt. These rays can penetrate into an ordinary UHMWPE implant without any problems. When these photons penetrate into the polyethylene, it will interact with an atom and put its electron(s) in motion, producing kinetic energy. This kinetic energy is called KERMA (Kinetic Energy Released in the Medium). (Premnath, 1742). The energy is then transferred to the polyethylene by breakage of bonds, excitation, and ionization. An orbital electron will be removed while forming a positively charged ion, a free radical, from the breakage of C-H or C-C bonds (refer to Figure 1). More precisely, the ejected electrons will be recaptured by the parent ions which form the free radicals. Unfortunately, these radicals are more mobile at room temperature therefore forming free radicals and reacting at the same time. (Premnath, 1746).
Figure 1 Figure 2
Free Radical Oxidized
This is known to be an instantaneous process occurring upon the sterilization. These free radicals will react with oxygen in the air and result in oxidation reactions (refer to Figure 2). Also, cross-linking and/or chain scission will occur depending on the environment in which it was sterilized (discussed later) and the dose of radiation.
In chain scission, long chains break down into shorter chains, allowing them to freely pack together, increasing its density. UHMWPE consists of highly entangled chains which is in a meta-stable form, but upon irradiation, these main chains are broken in the amorphous and interfacial regions, decreasing the density of the entanglement. This will result in the expansion of the crystalline region or the growth of crystalline lamella because polymer chains have a tendency to increase their crystalline region whenever there’s enough energy available. (Premnath, 1748). This will lead to lower molecular weight and higher % crystallinity. This higher crystallinity will harden the polymer, but it will increase its brittleness. Crystallinity can increase up to about 28% in 112 months. (Naidu, 141).
In cross-linking, the linear form of the chain will convert to a network structure, or a highly crosslinked structure, by bonding between the chains. The original linear chains will join each other by covalent bonds. Crosslinking can lead to an improvement in the mechanical properties such as wear and creep resistance. By controlling the amount of crosslinking or this formation of 3-D structure, beneficial properties of UHMWPE such as ductility and toughness can be retained. This technique can be carried out by exposing the polymer to an elevated temperature after gamma irradiation. This technique will be further discussed and compared to other methods of sterilization.
Figure 3
Gamma Radiation (photon emission)
(
Ejection of an Electron
(
Formation of Highly Reactive Free Radicals
(
Oxidation Reactions
(
(((
¯ (
Cross-Linking Chain Scission
(
Change in Mechanical Properties
(
Generation of Polyethylene Debris (Delamination)
(
Osteolysis, Loss of Bony Architecture
(
Loosening of the Implant, Failure
Oxidation reactions will proceed as oxygen diffuses into the polyethylene reacting with the free radicals. Amount of oxidation will depend on time; therefore oxidation will increase while it ages on the shelf and after implantation. The reason for the oxidation continuing after the implantation is due to the slow oxygen diffusion and the longevity of the radicals. Oxidation itself produces another form of free radical which in turn can react with oxygen. (Premnath, 1749). Due to the increase in supply of radicals, oxidation could occur depending on how much oxygen that are available for reaction. This oxidized surface of the polyethylene will grow in thickness as its aging. From past experiments, the maximum oxidation seems to occur at about 1 to 2 mm below the surface. (Technical Monograph). Oxidation will significantly alter some of the properties of the UHMWPE. The oxidation surface will become harder and more brittle compared to the core of the polymer. The embrittlement will decrease the ability of the material to elongate or lower its ductility. Since the toughness and the wear resistance is directly proportional to its ductility, they will decrease as well. (Morra, NO.227). This reduction in resistance to fatigue wear will in turn can cause pitting and delamination. In pitting, small pits or holes will form from very localized corrosion attacks or corrosion fatigue. Surface delamination or interfacial failure is one of the main cause of failure of total knee replacements by releasing wear debris. (Morra, 227). These fatigue failures seems to be caused by the decrease in toughness or increase in brittleness of the surface layer caused by oxidation. (Technical Monograph). The polyethylene wear debris are the main cause of osteolysis (refer to figure 4 and 5). Figure 5 shows the deformation and debris generation due to wear and friction of the UHMWPE. (Blanchard, Biomaterials:Body Parts of the Future).
Figure 4 Figure 5
[pic] [pic]
Unworn UHMWPE UHMWPE after friction
and wear test
20,000X
The debris initiates a macrophage response which involves a phagocytic tissue cell that protects the body against infections and harmful or destructive substances. The removal of these polyethylene wear particles are caused by the lymphatic channels in and around the artificial joints. This has been proven by the discovery of these POLYETHYLENE particles in the lymph nodes. The macrophage response near the tissue will cause osteolysis (death of bone cells), bone loss with enough time and loss of normal bony structure.(Morawski, 772-775). Finally, the implant will loosen resulting in failure.
Past experiments have proven that failure due to fatigue is caused by the increase in embrittlement of the oxidized surface. In one of the experiments, elongation was found to be 30 % lower on the surface than the core 53 months after sterilization by gamma radiation. (Technical Monograph). Both compression and tension fatigue resistance were lowered. In compression fatigue testing, crack propagation distance was double the length compared to a non sterile UHMWPE. In the tension fatigue testing, the rate of crack propagation was faster by an order of 2. (Goldman, 2912). Another study showed that there was a 26% increase in compressive modulus, a 16% increase in yield strength, an 11% decrease in elongation at break, and 20% decrease in the impact strength compared to the non-irradiated sample. (Sauer, 1932-1934). These data proves the increased brittleness of the polyethylene. Also, fatigue strength was 42% lower than that of the non-irradiated sample. (Sauer, 1932-1934). It is clear that gamma radiation reduces the resistance to fatigue and ultimately leads to the failure of the implant.
Gamma radiation initiates a series of reactions in the polyethylene which ultimately results in the failure of the implant. The free radicals created by the radiation will react with oxygen and oxidation will proceed. The production of free radicals seems to be the main factor that leads to the degradation of the UHMWPE. Depending on the environment in which the gamma radiation took place, the degree of oxidation, crosslinking, chain scission, and formation of free radicals will vary. Up to now, only radiation in an air environment was discussed, however, gamma radiation in an inert environment has different affects on the polyethylene. This and other techniques of radiation will be explained in the next section.
Differences Between Gamma Irradiation in Air and Inert Environment:
As previously discussed during the sterilization of UHMWPE by gamma radiation, chain scission and free radical formation occurs. This creates a condition of relative instability within the material whose eventual outcome can be significantly affected by the environment in which the UHMWPE finds itself. When irradiation takes place in air, the presence and high reactivity of oxygen begins rather aggressive oxidation of the UHMWPE. This leads to significant changes in the physical and structural properties of the material both during and subsequent to sterilization. In attempt to minimize this degradative oxidation, many manufacturers of polyethylene implants have turned to a sterilization process in which the components are gamma irradiated in an inert gas, such as argon or nitrogen. While this approach has led to a decrease in immediate oxidation, its long term benefits have recently been called in to question.
The changes in molecular structure, crystal morphology and mechanical properties of UHMWPE due to irradiation treatments have been studied since the 1970’s. With many discrepancies concerning the most important of mechanical properties as concerns longevity of arthoplastic implants – wear. (Premnath, 1741) Recently manufacturers have turned to irradiation sterilization of POLYETHYLENE in oxygen evacuated packaging (argon and nitrogen filled) to minimize the levels of degradative oxidation.
The results of several studies have speculated that “[sterilization of polyethylene with radiation in the absence of oxygen might diminish or eliminate [initial] oxidative degradation. Sterilization in nitrogen has been shown to diminish oxidative degradation during the sterilization process but, on subsequent exposure to oxygen, degradation resume[s].” (Rimnac, 1055). This belief is substantiated by the findings of Sun, Wang, Stark, and Dumbleton. Using Fourier Transform Infrared Spectroscopy, Sun et al found that subsequent to oxidation, air-irradiated components showed an oxidation index level of 0.02, while nitrogen irradiated samples had a corresponding value of less than 0.01. However, upon simulated aging, the air- and nitrogen-irradiated samples had oxidation index levels of 0.11 and 0.08, respectively. (Sun et al, 3) It has been speculated that over long periods of aging the oxidation levels in the two processes will converge to a common value. According to Ries et al (p. 87), in a different study, at a depth of 3.5 mm, the oxidation levels were equivalent for gamma radiation in air and in argon gas, 0.05.
This thinking and its substantiation are consistent with the theory of chain scission and free radical formation that occur as a result of the gamma irradiation. When the sample is irradiated in the absence of oxygen, free radicals can remain in the material for quite some time, perhaps several years. This is especially true of the free radicals which are present in the crystalline regions of the material where the rigid structure makes it difficult for chains to mobilize. As a result, the radicals in the crystalline region travel along the stems of the crystal lattice to the more reactive amorphous and interfacial regions. (Premnath, 1748) It is necessary then to consider not only the exposure to oxygen after the irradiation process but to consider also the length of exposure to oxygen prior to the sterilization treatment. Premnath et al calculated that it would take one hundred days for an oxygen front to travel 1 cm in a sample of POLYETHYLENE with a 52 percent crystallinity and typical experimental diffusivity values (p. 1748). Following this logic, a 2 cm sample that was allowed to equilibrate with the atmosphere for one hundred days or more would have high levels of internal oxygen which be available for post-irradiation oxidation, regardless of the oxygen levels in the surrounding atmosphere.
Despite these differences in the oxidation levels and the findings of a research group led by Gomez that high oxidation levels are concurrent with high levels of degradation, some feel that it is not yet possible to draw a direct conclusion between the level of oxidation and the longevity of the implant in vivo. This is in part due to the variability of several other factors including: date of sterilization, resin type, shelf life , length of implantation, patient weight and life style.(Gomez, 188-9)
Another physical property that is affected by gamma irradiation, and thus worthy of investigation is the extent to which cross-linking occurs in the material. “[W]hen POLYETHYLENE is irradiated in an inert environment, cross-linking, rather than chain scission and oxidative degradation dominates. The mechanical properties of POLYETHYLENE could be substantially affected, depending on the extent of cross-linking”(Rimnac, 1055). In the absence of oxygen, cross-linking in POLYETHYLENE dominates over chain scission. While there is sufficient energy present under typical sterilization irradiation levels to break the carbon-carbon bonds of the main chain, or backbone, without oxygen present the rupture of the carbon-hydrogen bonds is more likely. This leads to increased levels of cross-linking (Premnath, 1746). The ability of this increased cross-linking to increase the wear resistance of the UHMWPE is a topic that has been addressed by Dr. D.C. Sun et al.
In a recent study, this group has developed a process whereby free radicals are encouraged to exhibit cross-linking. According to their research, the process yields a POLYETHYLENE with improved mechanical properties. The results of Dr. Sun et al, as presented at the Sixty-Third Annual Meeting of the American Academy of Orthopaedic Surgeons, showed that components which had been sterilized under gamma radiation and subsequently stabilized maintained a low molecular weight fraction and were practically not subjected to aging-related oxidation. The unaged, stabilized UHMWPE contained a low molecular weight fraction of 18 percent, while the gamma irradiated samples in air had a mean low molecular weight fraction of 28 percent. A decrease in wear properties has been shown to be directly linked to this increased low molecular weight fraction. The difference in the low molecular weight fractions of these two sterilization treatments is even more pronounced in simulated aging testing. After a period of twenty three days in a temperature of 80 degrees Celsius, which was calculated to approximate eight years of shelf aging, the air-irradiated sample showed a low molecular weight fraction of 47%, and the stabilized samples averaged 8%. This data would suggest that prolonged shelf aging is particularly detrimental to the air-irradiated samples, while actually somewhat beneficial to stabilized components.
The smaller low molecular weight fractions of the stabilized samples correspond to increased wear resistance in the Sun group study. In an experiment designed to simulate in vivo conditions of the hip prosthetic, air-irradiated samples showed a wear rate of 27.8 milligrams/million cycles. The test was conducted by subjecting the samples to loads of up to 2450 Newton at a frequency of 1 Hertz on an MTS eight station hip joint wear simulator. The articulation took place in a bovine serum to approximate implant conditions. The same experiment conducted on the stabilized samples yielded a wear rate of 16.6 mg./million cycles. The simulated-aged air-irradiated samples lost 106.7 mg per million cycles under this testing, while the simulated-aged stabilized sample wear rate remained virtually unchanged, 16.1 mg./million cycles.
Based on these findings, it seems clear that gamma irradiation of POLYETHYLENE in air is clearly inappropriate and should be discontinued as a means of sterilization. This conclusion is corroborated by Premnath et al and others.
Ultra High Molecular Weight Polyethylene articulating surface components from total knee implants (refer to Picture 1) were obtained and tested to determine the effects of gamma irradiation under different environments. One sample (Miller/Galante Total Knee System Tibial Articular Surface: Zimmer, Warsaw, Indiana) was sterilized with gamma irradiation in atmospheric air, the other sample (Nexgen( Complete Knee Solution Articular Surface, Zimmer, Warsaw, Indiana) was sterilized after packaging in an air-tight, argon filled environment.
Since oxidation is shown to increase after irradiation for periods lasting up to several years after sterilization, it was sought to determine the morphological changes in the UHMWPE as a function of time. This was accomplished by weekly measurements of the heat of fusion and percent crystallinity of both samples over a period of three weeks. The heat of fusion was measured with the Perkin-Elmer Differential Scanning Calorimeter (Model DSC 7). Before testing the polyethylene each week the Calorimeter was calibrated using a sample of indium whose heat of fusion is precisely known (28.45 Joules/gram). The indium sample had a mass of 14.138 grams as measured using the Perkin-Elmer AD-4 Autobalance (a more detailed procedure that will be explained later in this section).
To determine how the different sterilization techniques affected the material properties of the samples, three point bending testing was conducted on relatively uniform samples machined from the given components. The research group decided to conduct three point bending testing because this test closely simulates the compressive forces that both the knee and hip replacement components are subject to. Due the brevity of the project and time constraint on access to the laboratory, high cycle wear/friction testing was not feasible.
The samples for mechanical testing were cut from the components by the University of Pennsylvania Machine Shop (Ed) with strict instructions to carry out the machining and an extremely slow rate of speed with very sharp instruments to minimize the effects of heat related changes in the material (refer to Drawing 1). Drawing 1 shows the four strips or samples taken from both inert and air irradiated tibial inserts. Drawing 1 also displays the edges where pieces were removed for DSC testing.
Picture 1
[pic]
Drawing 1
UHMWPE Tibial Articular Surface
[pic]
Because the machined samples were likely to have slight surface imperfections due to the machining process it was decided that a notch should be added to the test samples to control the location of failure. A notch was cut into the samples with a machine (constructed by Alex Radin, University of Pennsylvania) capable of making the incisions at the velocity specified by the ASTM for notching plastics, 0.25mm/min. The machine was set to deliver a notch depth of 30% of the original width (as it turned out later the actual notches were in excess of the measured amount). The samples were tested beyond their yield point in three point bending on the Instron Testing Machine. An X-Y Recorder plotted a load vs. displacement curve for each run.
Additionally, a hardness test was devised to compare the relative resistance of the differently treated components to local deformation. This test was conducted by pressing a specially treated super hard steel ball into the UHMWPE samples. The polyethylene was placed on top of a carbide plate on top of the lower Instron cross-head. A carbide plate was also placed between the steel ball and the upper cross-head. The carbide was used to insure that all of the applied load of the Instron was used in pressing the steel ball into the UHMWPE and not in deforming the cross head with such high localized pressure. Measurements of load and displacement were recorded and a hardness calculation similar to the Brinell hardness level was made. Due to the non-uniformity of the articular surface remaining after machining, all hardness testing was conducted on samples that had been machined.
Mechanical Testing:
Materials
• Instron Testing Machine Model 1331, Servohydraulic System (Stroke control, control
of the displacement: load is a function of displacement). Full capacity of the load cell - 2,000 lbs. with 5 ranges, 100%, 50%, 20%, 10%, and 5%.
• Ramp Test (Compression for 3 Point bending test).
• X-Y Recorder, Yokogaura Model 3023.
Operation of the Differential Scanning Calorimeter:
Materials
• Perkin-Elmer DSC7 Differential Scanning Calorimeter
• Perkin-Elmer Autobalance
• UHMWPE samples (Miller/Galante Total Knee System Tibial Articular Surface: Zimmer, Warsaw, Indiana(sterilized in air) and Nexgen( Complete Knee Solution Articular Surface, Zimmer, Warsaw, Indiana(sterilized in argon))
• Aluminum Pans and Lids
• Hermetical Sample Sealer Press
• Tweezers
• Digital Balance
Method
The calorimeter was calibrated each week prior to testing using a sample of Iridium between 100(C and 200(C at a rate of 40(C/min. Knowing the mass, melting temperature, and heat of fusion of the sample the calibration constant, Kc, was computed. Having completed the calibration of the calorimeter, the testing could begin in earnest.
Thin, round pieces of the UHMWPE tibial articular surface (refer to Drawing 1) was cut out from the outer edges. The pieces were cut out at the same location for both the inert and the air irradiated samples. Gloves were worn to ensure that no oil or any particles would get on the sample.
Each sample of UHMWPE was hermetically sealed in aluminum pans before testing. Before sealing, the difference in the mass of the pans was measured using he Perkin-Elmer Autobalance and recorded to provide a basis for error quantification. Additionally, the mass of the sample was measured and recorded.
The sealed sample was loaded into the left bed of the DSC7 furnace, while a reference pan was inserted into the right side of the furnace. Testing was conducted by heating the sample between the range of 50(C and 200(C at a temperature increment rate of 20(C. The computer recorded and plotted the data collected for all the runs.
To analyze the data, the areas under the plot of energy versus temperature were cut out of the printouts and weighed using a digital balance. Additionally, an area of known unit area was cut out and weighed. By equating the ratios of the areas and the masses, it was possible to determine the area under the curve. Cutting along the offset line, gave a weight for the crystalline portion of the sample, which was used to calculate the percent crystallinity of the sample.
In preparation for this project, it was utmost necessary to do research on the best methods of testing in order to produce data of consequence with these samples. In our tests and studies, we have learned that a means of determining the differences in mechanical properties of many structures is through the determination of the molecular structure. With our samples of ultra-high molecular weight polyethylene, it has already been determined that they had been prepared by different means of sterilization. But in addition to this, they had also been constructed differently as they also differed in size and shape. Therefore, our decision was to use the differential scanning calorimeter (DSC), by which we would be able to measure the change in enthalpy of a melting sample when compared to a reference tray. The differential scanning calorimeter produces a plot of the increasing energy versus the temperature of the sample, from which, a best fit line may be placed in order to determine various physical properties of the polyethylene implant.
Differential Scanning Calorimeter
However, first, before we can begin testing the polyethylene samples, it was necessary to do a calibration curve on the DSC, which was needed before every new day of testing. This is accomplished by using an indium sample against an empty reference tray. This is under the assumption that indium is near complete crystallinity, and therefore, all samples of indium will have the same (and known) heat of fusion. With this knowledge, by melting a sample of indium and plotting its best fit line on the resulting graph, we are able to have the computer determine the heat of fusion of the sample. Now, substituting this into equation (1), and also having found out the area under the curve, as given to us by the computer, we are able to calculate the calibration factor of the heat of fusion that is needed in subsequent determinations for the polyethylene samples. The
[pic]
determination of the heat of fusion of any melting transition is calculated by finding the area under the peak
We applied equation (1) to indium calibrations taken at the beginning of each day of testing over the three weeks, for which the data may be found in the appendices. Table 1-1, which can be found in Appendix A, is the differential calorimeter scanning calibration data from the first day of testing (March 26, 1997). It shows that the heat of fusion of the indium sample, which is taken from the graph, in addition to the mass of the
|mass sample |crystal area |DH(J/g) |kc |
|17.5 |78.29 |28.04 |6.27 |
|17.5 |81.76 |28.04 |6.00 |
[pic]
[pic]
indium sample, which we weighed and measured when preparing the sample, and finally, it shows the area of the crystal portion of the sample under the curve which is found by a method of cutting and weighing of paper portions of the graph.
For every graph that was produced by the computer of the increasing energy plotted against temperature, we made a second graph of the same information, however, of a smaller portion of the graph so that it would be more accurate in our placing of the best fit line. This gave what appears to be two sets of data for each sample, but rather, it is an average of one set of data but given in two different presentations. We decided that this would be a means to follow to give us an estimation of how close our data is relative to each other.
Now, after the determination of the calibration factor, we had to complete one more step before being able to determine the heat of fusion of the samples. With the printout of data from each sample which we melted, the printout of the graph that results from that set of data gives us what appears to be a slowly constantly increasing curve followed by a sharp increase into a peak which then gradually comes back down and results in another constant slope portion of the graph. By fitting a best fit line between the two linear portions of the graph, the computer then proceeds to fit in a straight line
[pic]
from the tip of the peak to intersect with the best fit line. The portion to the right of this curve will give us data that is a result of the crystalline portion of the sample, while the portion to the right is due to the amorphous properties of the implant piece (Figure 1-1).
By a method of cutting and weighing, we managed to determine the percent crystallinity of the sample, a physical property that affects the mechanical values of the polyethylene. Therefore, our determinations of the percent crystallinity is very important in the results of data in this experiment. The corresponding calculations for the percent crystallinity of the samples can be found in the corresponding appendices once again. However, Figure 1-2 shows a comparison of the percent crystallinity values for the air and the inert samples that was found over the three weeks of testing.
[pic]
Having determined the percent crystallinity of the samples taken each day of testing allowed us the opportunity to compare the values found between the sample irradiated in the presence of air and the implant irradiated in the presence of the inert gas. Whether or not a trend is present, however, is really affected by the factor that only three sets of data were only taken, due to time restrictions. However, the significance of this data presented is further elaborated upon in the discussions section.
As explained previously, the determination of the percent crystallinity is necessary to determine the heat of fusion of each sample of polyethylene taken, as the
heat of fusion varies as a result of this physical property. With the determination of the area under the curve of the crystalline portion of the sample measured, we are able to now place these values into a re-arranged format of equation (1), as seen in equation (2)
[pic]
The calibration factor, kc is tested for each day; As is the area under the curve of the sample that we are interested in, which, in this case, is the crystalline portion of the graph as seen from Figure 1-2; and finally, ms is the mass of the sample which we are testing, which we were able to measure while preparing the trays. However, we noticed that for this to be truly accurate, it would require that the calibration factor having been plotted on a graph with axes whose incremental increase were identical to the axes of the samples that we were comparing. Therefore, for each sample tested, we printed up two graphs of different axes. From there, we were required to apply the correct calibration factor of corresponding axes. The calculations and methods of calculating are shown in the appendices. Figure 2-1 is a graphical representation of the average heats of fusion of
[pic]
the samples taken from the air- and inert-irradiated samples, compared over the three
weeks of testing. However, it should be noted, though, on the third week of testing, we were presented with the remainders of the implants that had not been used in the previous differential scanning calorimeter testing and was remaining from the machined pieces that had been used in mechanical testing. With these pieces available, it was possible for us to determine if there was any significant difference in the percent crystallinity and therefore, the heat of fusion of the sample between its surface and the core. This, from our understanding of our research, would be as a result of the oxidation and the depth of
[pic]
penetration of the gamma irradiation. It has already been discussed the effects of the gamma irradiation and its role in the formation of free radicals in the long chains of the
ultra-high molecular weight polyethylene, thus resulting in the change of the physical and material properties of the implant material. Therefore, we decided to make a comparison of the heat of fusion data from the third week of data by comparing the samples taken from the core and samples taken from the surface for each implant piece, as seen in figure 2-2. However, we found that the confidence limit of the core samples were greater than 10% of the average value and there was associated with it a very high standard deviation. As a result, we came to the conclusion that the tests done on the core are insignificant in
terms of this experiment because we do not have a proper means of comparing the two sets of data with the little sample that we have been given.
Mechanical Testing
In our pursuit of testing, we discovered that available to us on campus at the Laboratory of Research for Structural Materials, LRSM, we had available to us quite a few mechanical testing machines. However, we were presented with a dilemma because of our limited samples which we had to share with another group doing a similar project. In addition to this problem, we were given two samples of different shape, size and contour. Therefore, many of the mechanical tests which are comparisons over two pieces would be futile because of the difference in dimensions of the pieces. Also, if we were to pursue tests, the pieces which we were working with were not of uniform thickness, therefore varying our data results and making calculations complex and beyond our scope. However, in our further investigations of what was available to us at LRSM, we
|Sample |Lower Limit |Upper Limit |Young's Modulus (MPa) |
| |(MPa) |(MPa) | |
|Air (1) |1053.6 |1146.9 |1100.2 |
|Air (2) |1026.8 |1131.5 |1079.1 |
|Inert (1) |1326.3 |1408.7 |1367.5 |
|Inert (2) |956.0 |1092.4 |1024.2 |
were able to discover that there is indeed a method of machining the two implants to produce several samples from each implant of almost similar dimensions and constant thickness without affecting the material and physical properties of our implant pieces. The result was then that we would perform three point bending and also a compression test on the samples. The result of the three point bending tests is the determination of the Young’s Modulus of the samples, as presented in Table 3-1.
In addition to the three point bending tests, the compression tests allowed us to be able to do a mock Brinell Hardness Value for the samples. These were plotted against the
increasing load and averaged for the air and inert samples, showing a trend of an
[pic]
increased hardness value for the air sample (Figure 3-1). The significance of this data will also be further elaborated upon in the discussion section.
However, these tests were done on only a limited supply of material and was done over a day. Therefore, there is no real significance in our question about how time affects the oxidation effect of gamma irradiated polyethylene samples, but however, it does show a potential trend of the result of the different types of gamma irradiation in the presence of different environments.
Three weeks of testing yielded interesting findings in the different effects of methods of sterilization on the physical and mechanical properties of Ultra High Molecular Weight Polyethylene. Differential scanning calorimetry showed that the heat of fusion and percent crystallinity of gamma irradiated polyethylene in air remained constant within the limits of statistical error over the three week time frame, while the polyethylene samples irradiated in an inert atmosphere showed a trend to increase in both the heat of fusion and percent crystallinity values over the same time period. Despite being able to note the aforementioned trends in the properties of UHMWPE, the actual values found for the percent crystallinity of both samples must be questioned. This is due to the fact that in the heat of fusion values, the inert samples showed an initially lower result than that of the air irradiated samples, and in the percent crystallinity calculations the air irradiated samples had a lower initial value. The heat of fusion values for the air-irradiated samples had an initial mean value of 121.98 J/g ( 4.0% while the inert-irradiated samples showed a mean heat of fusion for the first day of 104.47 J/g ( 3.0%. The percent crystallinity values of the air-irradiated samples were 72.65958 ( 2.0% at the first week of testing, and the inert-irradiated samples showed initial levels of crystallinity of 78.44 ( 1.0%. This is clearly contradictory. In fact, another group conducting the same experiment reported similar findings on their first day of data collection (one full day after our first day of collection). Group R2 group (Fernandes, Gupta, James and Kiok) found the air-irradiated samples to have a mean heat of fusion of 106.32 J/g (no deviation reported) and an average percent crystallinity of 81.57 (no reported deviation). The same group reported for air-irradiated samples values of heat of fusion at 99.82 J/g (no reported deviation) and mean percent crystallinity of 82.60% (no reported deviation). Again, the heat of fusion is high in the air-irradiated samples and percent crystallinity high in the inert-irradiated samples. The percent crystallinity and heat of fusion values should how consistency in relative values, i.e. if the heat of fusion is lower in one of the samples than the percent crystallinity of that sample should also be lower than that of its counterpart. This is due to the fact that the increased crystallinity results in a more organized packing structure than a more amorphous sample, therefore the molecular strand of the polyethylene can be brought in to loser contact, creating stronger intermolecular attractions. A possible exception to the necessity of these two characteristics following each other is the case of polyethylene samples having similar oxidation levels (and thus percent crystallinity) yet different levels of crosslinking. Such a situation should produce similar measurements for the level of crystallinity yet different heats of fusion. The sample with a higher degree of crosslinking would show increased heat of fusion as well as melting temperature.
The nature of the our testing was such that the percent crystallinity of the samples is a based on a calculation which further processes the findings of the heat of fusion values. The largest source of error inherent in measuring the percents crystallinity is the ability to manually cut out a portion of the computer generated graphs from the Perkin-Elmer DSC-7. In an attempt to quantify the error inherent in this process, five squares with five centimeter edges were cut out of the same paper stock used for the DSC printouts. The weights of each were determined and a deviance of 2% was calculated. This represents a greater percent of error present in either the DSC or the scale used to measure the cut-outs (Mettler Toledo PB303 DeltaRange). It should be obvious that even larger errors are to be expected in the cutting of shapes that are more complex and have larger perimeters than the simple squares used in the error quantification process, therefore even larger deviance is expected in the cut-outs of the graphs. It is because of this that in the matter of the two conflicting data sets, the percent crystallinity values were discredited and disregarded.
The nature of the observed trends is seen in the graphs presented in the results section. In both the heat of fusion and the percent crystallinity, the mean values of the air-irradiated samples fluctuated over the three week period, falling at the two week mark and then rising at the third week. For instance, the initial heat of fusion of the air-irradiated samples, as previously noted, was 121.98 J/g ( 4.0%. At the second week the heat of fusion was 111.21 J/g ± 2.5 %. By the third week, the mean heat of fusion had risen to 118.74 J/g ± 5.3%. The reported ranges of deviation indicate that some of these values may be statistically the same, and in deed, the paired sample t-test for means confirms this for all but the difference between the first and second weeks of testing. It is felt that further testing would have shown that the heat of fusion of the air-irradaiated sample would not deviate significantly from week to week. The percent crystallinity of the air-irradiated samples showed a similar trend: falling from 72.65 ± 2.0% the first week to 71.96 ± 1.7% at the second week, and then rising to 75.04 ± 2.2% on the third week. Again, paired sample t-tests for means were performed, showing that in the percent crystallinity, there was no significant difference between the means reported each of the three weeks. As stated previously, we have little confidence in the absolute values of the percent crystallinity values. We are confident, however, that they are useful, when combined with the heat of fusion data in showing the trend in the rate of oxidation in the air-irradiated samples, a basically stagnant (over short periods) value.
Similar analysis was conducted for the inert-irradiated samples. The heats of fusion measured each week for these samples showed a slight but consistent increase. The differences were so slight that in fact, a t-test conducted for the means between weeks one and three and between weeks one and two reported that the differences were not statistically different. It is the group’s consensus, however, that in light of the fairly steady increase observed from week to week these differences are in fact "significant” and continued study would more conclusively demonstrate this trend. The initial value of the heat of fusion for the inert-irradiated samples was 104.47 J/g ± 3.0%, the following week the mean heat of fusion had risen to 111.56 J/g ± 2.2%, and by the third week the mean heat of fusion was a high value of 125.12 J/g ± 10.6%. The percent crystallinity numbers reported a similar trend. The mean initial crystallinity measured 78.44 ± 1.0% at week one, week two was statistically the same with a mean of 78.39 ± 1.5%, and by week three the mean percent crystallinity had risen to 80.33 ± 1.9%. We again predict that prolonged study would show the levels of crystallinity to increase to some level which is dependent on the dosage of radiation exposure and the extent to which crosslinking occurs.
It should be noted here that on the second day of testing, despite the fact that the proper channels had been followed to obtain access to the DSC-7 during our normal laboratory hours, upon our arrival to the lab the DSC-7 was in use by another experimenter who was conducting experiments at -100(C. Despite showing extreme patience in waiting for the DSC to warm back up to approximately +100(C, the normal starting point for heat of fusion testing with UHMWPE, the data were still a bit flawed. The indium calibration procedure yielded a heat of fusion of 32.89 J/g. The standard reference value for the heat of fusion of Indium is given in most references as 28.42 J/g. This difference represents a 15.72% error over the reference value. It is unclear, whether or not this error can be simply added to the values as previously calculated and presented in the results section of this paper. Simply adding 15.72 percent to all day two data shows very unlikely trends for the calorimetric data in both the air and inert irradiated samples to increase over the time period between weeks one and two, and then to decrease over the final week interval. So while some correction factor may be in order, it is not at all clear what such a correction should be.
The mechanical testing was carried out over only one week and was limited to two samples for each sterilization method in three point bending, due to limitations of the quantity of material provided. The Young’s Modulus as determined from the three point bending testing was not significantly different between the two sterilization methods. The Young’s Modulus of the air-irradiated samples was found to be 1089.691 MPa with a standard deviation of 14.910 MPa, while the modulus of elasticity of the inert-irradiated samples was found to be 1195.87 MPa with a deviation of 242.76 MPa. A paired sample t-test for means showed that that there was no significant difference between the elastic modulus in the air- and inert-irradiated sterilized samples. Group R2 reported similar findings showing no significant difference in the Young’s Modulus of the differently sterilized samples. However a “Brinell-like” hardness test showed quite significant differences in the ability of the UHMWPE to resist local deformations. The inert irradiated samples permitted an average 22.56% greater penetration of the stainless steel ball into the material than the air-irradiated samples with a standard deviation of 2.82%. This shows that the air-irradiated samples do in fact have higher levels of oxidation (brittleness) and crystallinity, supporting our claim that the previously calculated values for the crystallinity must be disregarded.
Based on the data collected, we have been able to draw several hypotheses and conclusions. We feel that the evidence is convincing in showing that the rate of oxidation in UHMWPE post sterilization is an exponential decay function. This is evidenced by the noticeable changes in the physical properties of the inert-irradiated polyethylene almost immediately after initial exposure to oxygen (air), and the indiscernible changes of the air-irradiated samples upon the opening of the sterile packaging. The inert-irradiated sample is “empty” of oxygen and the potential for oxidation is initially high. The air-irradiated sample on the other hand (if it has, as we will assume, aged for any considerable length of time) will have oxidized to some extent. Thus, its potential to further oxidize is lower and the rate of oxidation slowed accordingly.
We are further convinced that the inert-irradiated sample must have received a higher dosage of the sterilizing gamma irradiation. This is due to the fact that despite its heat of fusion value that is initially lower than that of the air-irradiated sample, it eventually surpasses the heat of fusion of its counterpart. This indicates a higher initial overall potential for oxidation, which must be brought about by a greater production of free radicals during irradiation.
Additionally, we speculate that the inert-irradiated sample must have “seen” some oxygen prior to the opening of the “air-tight” packaging. This is thought to be so in light of the initial value of the inert sample’s heat of fusion being so close to that of the air-irradiated sample. Two mechanisms potentially explain the possibility for the exposure to oxygen. The most likely of these is the inability of the polyethylene which is used in the construction of the packaging to serve as an adequate long-term barrier to oxygen. Another source of oxygen is from within the material itself. Given the slightly porous nature of the material, it is likely that after the manufacture of the implant component, oxygen can be absorbed by the material. This oxygen then remains idle within the implant until conditions are favorable for reaction, i.e. gamma irradiation and the formation of free radicals.
In closing, we would like to make several recommendations. Firstly, it is our opinion, along with that of several other researchers in the field that the practice of UHMWPE sterilization by gamma radiation in air should be discontinued. This is due to the increased hardness shown under the mechanical testing, which will eventually translate into increased brittleness and possibly early failure. Additionally, we would like to recommend that alternative packaging be considered for those components that are sterilized in an inert environment. The plastics currently used show little promise as a long-term barrier to oxygen. To this end, systems must be considered that would guarantee timely delivery of implants component to those in need of them with a minimization of the shelf life of these components. Finally, the level of oxygen absorption between manufacture and packaging must be considered and minimized. Again, manufacturers must resist the temptation to churn out many more than needed at a time and then allowing these pieces to sit.
1. Bartley, R.E., Stulberg, S.D., Robb, III, W.J., and Sweeney, H.J.: Polyethylene Wear in Unicompartmental Knee Arthroplasty. Clinical Orthopaedics and Related Research. 299:18-24, 1994.
2. BE210 Bioengineering Laboratory II Laboratory Manual, Spring 1997.
3. Bhateja, S. K; Andrews, E. H.; and Yardbrough, S. M. Radiation Induced Crystallinity Changes in Linear Polyethylenes: Long Term Aging Effects. Polymer, 21:739-750, 1989.
4. Blanchard, Cheryl R. Biomaterials: Body Parts of the Future. .
5. Callister, William D. Materials Science and Engineering an Introduction, Copyright 1997, by John Wiley & Sons, Inc.
6. Collier, J.P., Sutula, L.C., Currier, B.H., Wooding, R.E., Williams, I.R., Farber, K.B., and Mayor, M.B.: Overview of polyethylene as a bearing material: comparison of sterilization methods. Clinical Orthopaedics and Related Research. 333:76-86, 1996.
7. Effects of sterilization methods on ultra-high molecular weight polyethylene. Technical Monograph.
8. Goldman, M.; Gronsky, R.; Ranganathan, R. and Pruitt, L. The effects of gamma radiation sterilization and ageing on the structure and morphology of medical grade ultra high molecular weight polyethylene. Polymer 1996, 37, 14.
9. Gomez-Barrena, Enrique; Chang, Jun-Dong; Li, Stephen; Rimnac, Clare M.; Salvati Eduardo A. The Role of Polyethylene Properties in Osteolysis After Total Hip Replacement. Instructional Course Lectures. Chapter 22.
10. Li, Stephen and Burstein, Albert H. Current Concepts Review. Ultra-High Molecular Weight Polyethylene. The Material and Its Use in Total Joint Implants. The Journal of Bone and Joint Surgery, Vol. 76-A(7), July 1994. Pp 1080-1090.
11. Morawski, David R.; Coutts, Richard D.; Handal, Edgar G. Polyethylene Debris in Lymph Nodes after a Total Hip Arthroplasty. The Journal of Bone and Joint Surgery, Vol. 77-A, NO. 5, May 1995.
12. Morra, Edward A. American Academy of Orthopaedic Surgeons 1996 Annual Meeting - Scientific Program. Effects of Plateau Geometry on Delamination Wear Failures in UHMWPE Tibial Inserts. Paper No. 227.
13. Naidu, Sanjiv H.; Bixler, Brian L.; Moulton, Mark J.R. Radiation-Induced Physical Changes in UHMWPE Implant Components. Basic Science 1996 February, Vol. 20, No 2.
14. Premnath, V; Harris, W.H.; Jasty, M.; Merril, E.W. Gamma sterilization of UHMWPE articular implants: an analysis of the oxidation problem. Biomaterials, 1996, Vol. 17 No. 18.
15. Ries, M.D., Weaver, K., Rose, R.M., Gunther, J., Sauer, W., and Beals, N.: Fatigue strength of polyethylene after sterilization by gamma irradiation or ethylene oxide. Clinical Orthopaedics and Related Research. 333:87-95, 1996.
16. Rimnac, Clare M.; Klein, Robert W.; Betts, Foster. Post-Irradiation Aging of Ultra-High Molecular Weight Polyethylene. The Journal of Bone and Joint Surgery. Incorporated, Vol. 76-A, NO. 7, July 1994.
17. Roe, R. J.; Grood, E. S. Effect of Radiation Sterilization and Aging on Ultra High Molecular Weight Polythrylene. Journal of Biomedical Materials Research, 15(2):209-30, March 1981.
18. Sauer, Willard L.; Weaver, Kevin D.; Beals, Neil B. Fatigue Performance of ultra-high-molecular weight polyethylene: effect of gamma radiation sterilization. Biomaterials, 1996, Vol. 17 NO. 20.
19. Shen, C and Dumbleton, J. H. The Friction and Wear Behavior of Irradiated Polyethylene. Wear, 30: 349-364, 1974.
20. Sun, D.C.; Wang, A.; Stark C.; Dumbleton J.H. Development of Stabilized UHMWPE Implants with Improved Oxidation Resistance via Crosslinking. Scientific Exhibition Presented at the 63rd Annual Meeting of the American Academy of Orthopaedic Surgeons, February 22-26, 1996, Atlanta, Georgia, USA.
21. Sutula, Lauren C.; Collier, John P.; Saum, Kenneth A. Impact of Gamma Sterilization on Clinical Performance of Polyethylene in the Hip. Clinical Orthopaedics and Related Research, Number 319, October, 1995.
22. Wasielewski, R.C., Jorge, O.G., Leighty, R.M., Raghu, N.N., and Rosenberg, A.G.: Wear patterns on retrieved polyethylene tibial inserts and their relationship to technical considerations during total knee arthroplasty. Clinical Orthopaedics and Related Research. 299:31-43, 1994.
23. White, S., Paxson, R.D., Tanner, M.G., and Whiteside, L.A.: Effects of sterilization on wear in total knee arthroplasty. Clinical Orthopaedics and Related Research. 331:164-74, 1996.
| |square mass |square |total mass |crystal |crystal area |total area |% crystal |
| | |area | |mass | | |- linity |
| |35 |20 |139 |137 |78.29 |79.43 |98.56 |
| |34 |20 |140 |139 |81.76 |82.35 |99.29 |
|Average : |34.5 |20 |139.5 |138 |80.03 | |98.92 |
[pic]
|mass sample |crystal area |DH(J/g) |kc |
|17.5 |78.29 |28.04 |6.27 |
|17.5 |81.76 |28.04 |6.00 |
[pic]
[pic]
[pic]
|sample |square |square |total |crystal |Crystal |Total |% Crystal |
|number |area |mass |mass |mass |area |area |- linity |
| | | | | | | | |
|1a |50 |37 |115 |82 |110.81 |155.41 |71.30 |
|1b |25 |27 |164 |117 |108.33 |151.85 |71.34 |
|2a |20 |23 |115 |86 |74.78 |100.00 |74.78 |
|2b |10 |16 |154 |115 |71.88 |96.25 |74.68 |
|3a |50 |42 |135 |100 |119.05 |160.71 |74.07 |
|3b |25 |27 |177 |133 |123.15 |163.89 |75.14 |
|4a |20 |17 |99 |68 |80.00 |116.47 |68.69 |
|4b |10 |18 |181 |129 |71.67 |100.56 |71.27 |
| | | | | | | | |
| | | | | | |Average |72.66 |
|Average |72.66 |
|n |8.00 |
|t-value |2.36 |
|standard deviation |2.33 |
| | |
|confidence limit |1.95 |
[pic]
[pic]
[pic]
|Sample # |kc |Crystalline |Mass of |DH(J/g) |
| | |Area |sample | |
| | | | | |
|1a |6.17 |110.81 |5.59 |122.19 |
|1b |6.00 |108.33 |5.59 |116.23 |
|2a |6.17 |74.78 |3.56 |129.41 |
|2b |6.00 |71.88 |3.56 |121.02 |
|3a |6.17 |119.05 |6.00 |122.39 |
|3b |6.00 |123.15 |6.00 |123.19 |
|4a |6.17 |80.00 |3.83 |128.99 |
|4b |6.00 |71.67 |3.83 |112.44 |
| | | | | |
| | |Average | |121.98 |
|average |121.98 |
|n |8.00 |
|t-value |2.36 |
|standard deviation |5.74 |
| | |
|confidence limit |4.80 |
|sample |square |square |total |crystal |crystal |total |% Crystal |
|number |area |mass |mass |mass |area |area |- linity |
| | | | | | | | |
|5a |50 |29 |142 |109 |187.93 |244.83 |76.76 |
|5b |25 |23 |204 |161 |175.00 |221.74 |78.92 |
|6a |50 |40 |150 |120 |150.00 |187.50 |80.00 |
|6b |25 |30 |225 |178 |148.33 |187.50 |79.11 |
|7a |50 |32 |134 |101 |157.81 |209.38 |75.37 |
|7b |25 |28 |218 |171 |152.68 |194.64 |78.44 |
|8a |50 |36 |114 |89 |123.61 |158.33 |78.07 |
|8b |25 |25 |148 |114 |114.00 |148.00 |77.03 |
|9a |50 |30 |111 |88 |146.67 |185.00 |79.28 |
|9b |25 |21 |152 |120 |142.86 |180.95 |78.95 |
|10a |20 |20 |147 |114 |114.00 |147.00 |77.55 |
|10b |10 |14 |196 |157 |112.14 |140.00 |80.10 |
|11a |20 |21 |115 |92 |87.62 |109.52 |80.00 |
|11b |10 |14 |164 |129 |92.14 |117.14 |78.66 |
| | | | | | | | |
| | | | | | |average : |78.45 |
|Average |78.45 |
|n |14.00 |
|t-value |2.16 |
|standard deviation |1.37 |
| | |
|Confidence Limit |0.79 |
[pic]
[pic]
[pic]
|Sample # |kc |Crystalline |Mass of |DH(J/g) |
| | |Area |sample | |
| | | | | |
|5a |6.17 |187.93 |11.14 |104.00 |
|5b |6.00 |175.00 |11.14 |94.24 |
|6a |6.17 |150.00 |8.73 |105.95 |
|6b |6.00 |148.33 |8.73 |101.95 |
|7a |6.17 |157.81 |9.03 |107.84 |
|7b |6.00 |152.68 |9.03 |101.52 |
|8a |6.17 |123.61 |7.03 |108.43 |
|8b |6.00 |114.00 |7.03 |97.30 |
|9a |6.17 |146.67 |7.85 |115.27 |
|9b |6.00 |142.86 |7.85 |109.25 |
|10a |6.17 |114.00 |6.75 |104.22 |
|10b |6.00 |112.14 |6.75 |99.76 |
|11a |6.17 |87.62 |5.14 |105.23 |
|11b |6.00 |92.14 |5.14 |107.68 |
| | | | | |
| | | |average : |104.47 |
|Average |104.47 |
|n |14.00 |
|t-value |2.16 |
|standard deviation |5.36 |
| | |
|confidence limit |3.09 |
|graph |square |square |crystal |crystal |
| |mass |area |mass |area |
|a |22 |50 |35 |79.55 |
|b |38 |50 |60 |78.95 |
| | | | | |
| | | |average: |79.25 |
[pic]
|mass |DHf |Crystal area |kc |
|14.14 |32.60 |79.55 |5.79 |
|14.14 |33.17 |78.95 |5.94 |
[pic]
[pic]
|Sample # |Square |Square |Total |Crystal |Crystal |Total |% Crystal |
| |Area |Mass |Mass |Mass |Area |Area |- linity |
| | | | | | | | |
|1a |50 |38 |103 |76 |100.00 |135.53 |73.79 |
|1b |25 |30 |147 |108 |90.00 |122.50 |73.47 |
|2a |40 |41 |122 |88 |85.85 |119.02 |72.13 |
|2b |20 |31 |177 |126 |81.29 |114.19 |71.19 |
|3a |50 |45 |108 |77 |85.56 |120.00 |71.30 |
|3b |25 |37 |161 |118 |79.73 |108.78 |73.29 |
|4a |40 |47 |159 |112 |95.32 |135.32 |70.44 |
|4b |40 |65 |217 |152 |93.54 |133.54 |70.05 |
| | | | | | | | |
| | | | | | |average: |71.96 |
|Average |71.96 |
|n |8.00 |
|t-value |2.36 |
|standard deviation |1.44 |
|confidence limit |1.20 |
[pic]
[pic]
[pic]
|Sample # |kc |Crystalline |Mass of |DH(J/g) |
| | |Area |sample | |
| | | | | |
|1a |5.79 |100.00 |5.12 |113.09 |
|1b |5.94 |90.00 |5.12 |104.36 |
|2a |5.79 |85.85 |4.35 |114.32 |
|2b |5.94 |81.29 |4.35 |110.99 |
|3a |5.79 |85.56 |4.36 |113.61 |
|3b |5.94 |79.73 |4.36 |108.56 |
|4a |5.79 |95.32 |4.93 |112.04 |
|4b |5.94 |93.54 |4.93 |112.73 |
| | | | | |
| | | |average : |111.21 |
|Average |111.21 |
|n |8.00 |
|t-value |2.36 |
|standard deviation |3.30 |
| | |
|confidence limit |2.76 |
|sample # |square |square |total |crystal |crystal |total |% crystal |
| |area |mass |mass |mass |area |area |- linity |
| | | | | | | | |
|5a |40 |48 |116 |93 |77.50 |96.67 |80.17 |
|5b |40 |69 |164 |130 |75.36 |95.07 |79.27 |
|6a |40 |62 |112 |88 |56.77 |72.26 |78.57 |
|6b |40 |90 |172 |133 |59.11 |76.44 |77.33 |
|7a |40 |52 |102 |81 |62.31 |78.46 |79.41 |
|7b |40 |77 |145 |117 |60.78 |75.32 |80.69 |
|8a |40 |70 |111 |85 |48.57 |63.43 |76.58 |
|8b |40 |97 |155 |116 |47.84 |63.92 |74.84 |
|9a |20 |46 |180 |134 |58.26 |78.26 |74.44 |
|9b |20 |56 |218 |167 |59.64 |77.86 |76.61 |
|10a |40 |73 |106 |86 |47.12 |58.08 |81.13 |
|10b |40 |91 |139 |110 |48.35 |61.10 |79.14 |
|11a |40 |58 |119 |95 |65.52 |82.07 |79.83 |
|11b |40 |79 |151 |120 |60.76 |76.46 |79.47 |
| | | | | | | | |
| | | | | | |average : |78.39 |
|Average |78.39 |
|n |14.00 |
|t-value |2.16 |
|standard deviation |2.10 |
| | |
|confidence limit |1.21 |
[pic]
[pic]
[pic]
|sample # |kc |crystalline |mass of |Dh(j/g) |
| | |area |sample | |
| | | | | |
|5a |5.79 |77.50 |3.96 |113.53 |
|5b |5.94 |75.36 |3.96 |113.20 |
|6a |5.79 |56.77 |3.15 |104.39 |
|6b |5.94 |59.11 |3.15 |111.44 |
|7a |5.79 |62.31 |3.20 |112.67 |
|7b |5.94 |60.78 |3.20 |112.69 |
|8a |5.79 |48.57 |2.61 |107.90 |
|8b |5.94 |47.84 |2.61 |108.96 |
|9a |5.79 |58.26 |3.12 |108.33 |
|9b |5.94 |59.64 |3.12 |113.71 |
|10a |5.79 |47.12 |2.36 |115.49 |
|10b |5.94 |48.35 |2.36 |121.50 |
|11a |5.79 |65.52 |3.40 |111.74 |
|11b |5.94 |60.76 |3.40 |106.25 |
| | | | | |
| | | |average |111.56 |
|Average |111.56 |
|n |14.00 |
|t-value |2.16 |
|standard deviation |4.28 |
| | |
|confidence limit |2.47 |
|square |square |total |crystal |crystal |
|mass |area |mass |mass |area |
|19 |50 |25 |25 |65.79 |
|19 |25 |46 |46 |60.53 |
| | | |average |63.16 |
[pic]
|mass |crystal |DHf |kc |
|sample |area | | |
|14.14 |65.79 |28.37 |6.10 |
|14.14 |60.53 |28.93 |6.76 |
[pic]
[pic]
|sample # |square |square |total |crystal |crystal |total |% Crystal |
| |area |mass |mass |mass |area |area |- linity |
|Surface Sample | | | | | | | |
|1a |50 |23 |155 |120 |260.87 |336.96 |77.42 |
|1b |25 |21 |213 |164 |195.24 |253.57 |77.00 |
|2a |50 |21 |133 |99 |235.71 |316.67 |74.44 |
|2b |25 |20 |190 |143 |178.75 |237.50 |75.26 |
|3a |50 |26 |134 |94 |180.77 |257.69 |70.15 |
|3b |25 |18 |189 |148 |205.56 |262.50 |78.31 |
| | | | | | |average |75.43 |
|Core Sample | | | | | | | |
|4a |50 |26 |128 |94 |180.77 |246.15 |73.44 |
|4b |25 |19 |166 |127 |167.11 |218.42 |76.51 |
|5a |50 |29 |150 |109 |187.93 |258.62 |72.67 |
|5b |25 |20 |201 |144 |180.00 |251.25 |71.64 |
|6a |50 |25 |157 |121 |242.00 |314.00 |77.07 |
|6b |50 |34 |217 |167 |245.59 |319.12 |76.96 |
| | | | | | |average |74.71 |
|Average |75.43 |
|n |6.00 |
|t-value |2.57 |
|standard deviation |2.95 |
| | |
|confidence limit |3.10 |
|Average |74.71 |
|n |6.00 |
|t-value |2.57 |
|standard deviation |2.41 |
| | |
|confidence limit |2.53 |
[pic]
[pic]
[pic]
|Sample # |kc |Crystalline |Mass of |DH(J/g) |
| | |Area |sample | |
|Surface Sample | | | | |
|1a |6.10 |260.87 |13.00 |122.39 |
|1b |6.76 |195.24 |13.00 |101.53 |
|2a |6.10 |235.71 |11.28 |127.40 |
|2b |6.76 |178.75 |11.28 |107.09 |
|3a |6.10 |180.77 |10.42 |105.75 |
|3b |6.76 |205.56 |10.42 |133.28 |
| | | |average |116.24 |
|Core Sample | | | | |
|4a |6.10 |180.77 |9.08 |121.34 |
|4b |6.76 |167.11 |9.28 |121.64 |
|5a |6.10 |187.93 |9.84 |116.39 |
|5b |6.76 |180.00 |9.84 |123.56 |
|6a |6.10 |242.00 |12.82 |115.09 |
|6b |6.76 |245.59 |12.82 |129.45 |
| | | |average |121.25 |
|Average |116.24 |
|n |6.00 |
|t-value |2.57 |
|standard deviation |13.14 |
| | |
|confidence limit |13.79 |
|Average |121.25 |
|n |6.00 |
|t-value |2.57 |
|standard deviation |5.18 |
| | |
|confidence limit |5.44 |
|sample # |square |square |total |crystal |crystal |total |% crystal |
| |area |mass |mass |mass |area |area |- linity |
|Surface | | | | | | | |
|Sample | | | | | | | |
|7a |50 |28 |161 |124 |221.43 |287.50 |77.02 |
|7b |25 |18 |207 |163 |226.39 |287.50 |78.74 |
|8a |50 |27 |155 |128 |237.04 |287.04 |82.58 |
|8b |25 |21 |230 |193 |229.76 |273.81 |83.91 |
|9a |50 |33 |146 |116 |175.76 |221.21 |79.45 |
|9b |25 |21 |181 |143 |170.24 |215.48 |79.01 |
| | | | | | |average |80.12 |
|Core | | | | | | | |
|Sample | | | | | | | |
|10a |50 |29 |169 |143 |246.55 |291.38 |84.62 |
|10b |50 |40 |252 |208 |260.00 |315.00 |82.54 |
|11a |50 |35 |169 |135 |192.86 |241.43 |79.88 |
|11b |50 |46 |220 |172 |186.96 |239.13 |78.18 |
|12a |50 |36 |216 |170 |236.11 |300.00 |78.70 |
|12b |50 |45 |271 |215 |238.89 |301.11 |79.34 |
| | | | | | |average |80.54 |
|Average |80.12 |
|n |6.00 |
|t-value |2.57 |
|standard deviation |2.59 |
| | |
|confidence limit |2.72 |
|Average |80.54 |
|n |6.00 |
|t-value |2.57 |
|standard deviation |2.51 |
| | |
|confidence limit |2.63 |
[pic]
[pic]
[pic]
|Sample # |kc |Crystalline |Mass of |DH(J/g) |
| | |Area |sample | |
|Surface | | | | |
|Sample | | | | |
|7a |6.10 |221.43 |12.15 |111.16 |
|7b |6.76 |226.39 |12.15 |125.96 |
|8a |6.10 |237.04 |11.98 |120.63 |
|8b |6.76 |229.76 |11.98 |129.60 |
|9a |6.10 |175.76 |9.24 |115.97 |
|9b |6.76 |170.24 |9.24 |124.50 |
| | | |average |121.31 |
|Core | | | | |
|Sample | | | | |
|10a |6.10 |246.55 |12.99 |115.76 |
|10b |6.76 |260.00 |12.99 |135.31 |
|11a |6.10 |192.86 |12.32 |95.46 |
|11b |6.76 |186.96 |12.32 |102.57 |
|12a |6.10 |236.11 |9.41 |152.96 |
|12b |6.76 |238.89 |9.41 |171.53 |
| | | |average |128.93 |
|Average |121.31 |
|n |6.00 |
|t-value |2.57 |
|standard deviation |6.82 |
| | |
|confidence limit |7.16 |
|Average |128.93 |
|n |6.00 |
|t-value |2.57 |
|standard deviation |29.74 |
| | |
|confidence limit |31.21 |
-----------------------
Table of Contents
Abstract
Background
Table 1
C
C
C
C
C
C
H
C
O
C
H
C
H
C
H
C
C
C
C
C
C
C
H
C
H
C
H
C
H
C
+
C
H
C
Methods and Materials
Results
Equation (1)
Table 1-1
Figure 1-1
Figure 1-2
Equation (2)
Figure 2-1
Figure 2-2
Table 3-1
Figure 3-1
Discussion
References
Appendix
Appendix A
Indium Calibration Curve : 3/26/97
% Crystallinity Calculations : Air Samples (3/26/97)
Error Analysis
Heat of Fusion Calculations : Air Samples (3/26/97)
Error Analysis
% Crystallinity Calculations : Inert Samples (3/26/97)
Error Analysis
Heat of Fusion Calculations : Air Samples (3/26/97)
Error Analysis
Indium Calibration Curve : 4/9/97
Appendix B
% Crystallinity Calculations : Air Samples (4/9/97)
Heat of Fusion Calculations : Air Samples (3/26/97)
Error Analysis
Error Analysis
% Crystallinity Calculations : Inert Samples (4/9/97)
Error Analysis
Heat of Fusion Calculations : Inert Samples (4/9/97)
Error Analysis
Appendix C
Indium Calibration Curve : 4/16/97
Note : on this day of testing (the final day) we were presented with the final pieces from our samples. This gave us the opportunity to test from the core and surface of the implants, allowing us to see if there are any significant differences between the two sections of the same implant. Therefore, please note the appearance of what seems like two sets of data for the % crystallinity and heat of fusion for each the inert- and the air- irradiated samples. This is not so, but rather, it is the sets of data separated only by the location from which the sample that was tested was taken from.
% Crystallinity Calculations : Air Samples (4/16/97)
Error Analysis
Error Analysis
Heat of Fusion Calculations : Air Samples (4/16/97)
Error Analysis
Error Analysis
% Crystallinity Calculations : Inert Samples (4/16/97)
Error Analysis
Error Analysis
Heat of Fusion Calculations : Inert Samples (4/16/97)
Error Analysis
Error Analysis
................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.
Related searches
- free christmas computer background wall
- country christmas background images
- background of plastic surgery
- background on starbucks
- christmas background wallpaper
- free christmas background images
- background information on plastic
- background study of plastics
- christmas background images free download
- microsoft free desktop background downl
- illinois health care worker background check act
- historical background of special education