How Processing Conditions Can Cause Degradation and Failure in Medical Devices.
Some processors in the plastics industry have the mistaken notion that failures are "accidents" and little can be done to avoid them. However, detailed inquiries have revealed that nearly all failures can be prevented, if the causes and mechanisms are sufficiently understood and appropriate countermeasures taken.
Injection-molded medical devices have much in common with other injection molding applications in terms of material and process requirements. But they also differ in significant ways. After the component is molded, it is often subjected to a combination of one or more washing or cleaning steps, drying at elevated temperatures, and assembly employing one of several possible techniques--in order to produce a final product. Before the device leaves the plant, a terminal sterilization with ethylene oxide or with radiation is applied. In addition, medical devices, like other products, experience shipping, warehousing, and shelf-life conditions under various temperatures and humidity situations before the user ever touches the product. To ensure failure-free operations, one must anticipate how all the above conditions interact with the molded component. In this article, we focus primarily on injection-molded medical devices.
Techniques used in this study include ASTM D3895-92 [isothermal OIT (oxidative induction time)], employing a DuPont 1090 thermal analyzer with a 910 differential scanning calorimetry (DSC) cell, and ASTM D1925 (yellowness index) measured by a Hunter colorimeter. Stabilizer chemical concentrations were determined by high-performance liquid chromatography (HPLC) using established calibration systems.
A TA Instruments thermal analyzer 2100 with a 2910 differential scanning calorimeter cell and a liquid nitrogen cooling accessory calibrated with indium standards were used to monitor the thermal behavior of the resin and molded parts. Heating rates of 20[degrees]C/min and cooling rates of 10[degrees]C/min were used throughout the study.
A Rheometrics RFR fluid rheometer was used to characterize the melt in the paraliel plate configuration using 2.50 cm plates from 0.1 to 300 [rad.sup.-1] with the dynamic mode. In order to minimize sample degradation, a very rapid heating program that brings the sample to an equilibrium test temperature in less than 2 min was adopted.
From the dynamic viscosity vs. frequency plot, a Newtonian viscosity was calculated by graphically averaging the low shear data points and extrapolating to zero shear rate. To ensure temperature reproducibility, the sample temperature thermocouple was calibrated using the indium standard periodically. A Mitsubishi model CA-06 micro Karl Fisher moisture analyzer was used for determination of moisture levels.
Surface failures were characterized by scanning electron microscopy performed on either a JEOL 35CF or a JEOL FE-6300 field emission instrument after a palladium sputter coating for conductivity was applied.
Optimization of Injection Molding Processing Parameters
In order to optimize injection molding processing parameters, all conceivable variables associated with the injection molding machine and resin were taken into account:
* Material type and mold design.
* Mold and machine capacity match.
* Material conditioning.
* Flaw, knit line, and notch elimination.
* Identification of degradation pathway.
* Validation of processing parameters.
Often, the importance of the material type is overlooked, leading to excessive molded-in stress or improperly fused knit lines. Mold shrinkage behavior, a property primarily governed by the degree of crystallinity, is of the utmost importance in mold and cycle design.
Economics at times dictates mold rotation or mold transfer between machines within a given facility or between facilities. At times, shot capacity and molding machine barrel capacity are inadvertently allowed to become unbalanced. This could result in excessive melt residence time, which--combined with thermally sensitive materials--can lead to serious thermo-oxidative and molecular weight degradation that causes subsequent failures. In a previous study,  an amorphous polyethylene terephthalate copolymer (PETG) was found to be extremely sensitive to thermal degradation when the melt temperature exceeded about 230[degrees]C. At lower temperatures, the material demonstrated sensitivity to mechanical shear degradation. These degradation pathways alter the molecular weight of the molded parts as reflected in the melt viscosities.
Another material requirement that needs particular attention is conditioning, especially drying. Many of the engineering thermoplastics belong to a polymer family derived from step growth polymerization, and more specifically, from condensation polymerization. In these types of polymerization reactions, a low molecular weight species, frequently water or an alcohol, is eliminated for each mainchain linkage. However, this condensation reaction is also reversible. At the high melt temperatures in the injection molding press, when moisture is present, each available water molecule can cleave a main chain bond, leading to significant reductions in molecular weight. This is one of the most frequently encountered failure modes in engineering thermoplastics. In Fig. 1, which is adapted from Ref. 2, previously dried polycarbonate resin was exposed to ambient moisture conditions for various times before molding, and the Izod impact properties of the molded samples were measured against conditioning time. It can be se en that in about 10 hr of exposure, sufficient moisture has entered the material to cause a drastic reduction in the impact properties of the molded components.
The significance of the molecular weight to the product performance is shown in Fig. 2, with molecular weight represented by viscosity. In this case, failures are detected by cracking of the molded component after downstream processing. Evidently, molecular weight degradation has shifted the material into a regime where the tensile strength/molecular weight dependence is nearly exponential. This is consistent with the widely accepted  functional dependence of ultimate strength on molecular weight:
[[sigma].sub.u] = [[sigma].sub.[infinity]] - B/Mn (1)
where [[sigma].sub.u] is ultimate strength; [[sigma].sub.[infinity]], is strength at infinite molecular weight; B is a material constant; and Mn is number average molecular weight.
In fact, the strength dependence on molecular weight was found to be an even stronger function for polymethylmethacrylate (PMMA) (Fig. 3).
The very strong dependence in this particular instance was corroborated in several other well-documented studies, making a compelling case for the control of injection molding variables for the prevention of failures. Also, this strong dependence highlights the power of choosing key molecular properties as independent variables in the study. From the functional dependence of failure occurrence on measurable properties, fundamental insights on component performance can be established. In addition, once the molecular weight best suited for product performance is identified by melt viscosity, it can be converted to a simple melt flow rate number easily determined on a melt indexer. This allows a very easily implemented technique for quality control.
Moisture Reduction in Injection Molding
Because of the reversibility of the polymerization reaction, it is a well-established requirement in the plastics industry that all condensation polymers be thoroughly dried before melt processing. Therefore, for every molecule of water remaining in the polymer, potentially one main-chain bond could be hydrolyzed, thus leading to serious molecular weight reduction and property loss.
Equation 2 clearly establishes that the zero-shear viscosity, [[eta].sub.0], is related to the weight-average molecular weight:
[[eta].sub.0] = K Mw [3,4] (2)
where [[eta].sub.0] is steady shear (Newtonian) viscosity; and Mw is weight average molecular weight.
The correlation was found to be universally obeyed for linear polymers, based on both theoretical and experimental considerations. From the above relationship, it is observed that the steady shear viscosity is a very sensitive measure of molecular weight because of the 3.4 power. Even for the case of a two-component alloy, the viscosity was found to be a sensitive indicator of the relative molecular weight degradations at each step of the process. Since most GPC determinations use polystyrene as standards, the absolute molecular weight can be obtained only by extensive calibrations. As a result, a 5% accuracy over the long term is considered excellent for GPC. However, from Eq 2, more than a 15% variation in viscosity would result, an easily achieved experimental precision.
Finally, the rheological parameters were correlated with actual product performance to establish a master "failure envelope" for acceptance criteria. An ASTM melt flow test was used, which allowed multiple sites along the product flow to monitor process conditions in order to attain true "zero-defect" performance in the final products.
PBT Pre-Processing Drying
In the medical industry, components such as clamps, connectors, and bearing housings are machined from PBT because of its unique physical properties--low coefficient of friction, hardness, and high-temperature usage capability. The processor purchases PBT resin from the material supplier and extrudes it into rods, which are usually annealed at 300[degrees]F (148.9[degrees]C) to 400[degrees]F (204[degrees]C) for about 6 hr in an oil bath to remove the internal stress. The purpose of annealing is to control the dimensional stability of the machined part.
In one of the experiments, a rod was cut to a disk approximately 0.25 inch (6.35 mm) thick and then vacuum oven dried at 120[degrees]C for about 15 hr prior to the rheology test. A DSC sample was taken from the skin of the rod. It is evident from the Table that the material was degraded during extrusion. Hydrolysis is the most likely cause, although extrusion conditions could be a secondary factor.
Polyester Elastomer Melt Processing
Polyether-ester thermoplastic elastomers (TPE) based on polybutylene terephthalate (PBT) hard segment and tetramethylene ether (PTMO) soft segments constitute an important class of medical elastomers because of their wide property range, solvent bonding capability, oxidative stability, and processing ease. In a series of experiments, film samples of 40D hardness were prepared on a 38-mm laboratory extruder under similar processing conditions but at varying thicknesses. Figure 4 presents the OIT activation enthalpy plot of two of the films. The 100-micron (4-mil) and the 200-micron (8-mil) film exhibited identical activation enthalpies, indicative of identical chemistry. A two-layer composite of the thinner film also exhibited OIT identical to that of the thinner film, indicating that the thickness of the OIT sample is not a primary variable for the induction times measured. However, between the two sets of data, there was about a three-fold difference. An inquiry into the details of the film extrusion process revealed that the thinner film was processed at a slightly higher temperature (193[degrees]C vs. 177[degree]C). However, temperature alone could not account for the magnitude of the observed OIT discrepancy. It was concluded that the residence time for processing the thinner film must have been significantly longer than the time for the 200-[micro]m film.
Influence of Radiation Processing on Material Degradation
Radiation sterilization is becoming popular in the medical device and packaging industry because of its convenience and low cost. Other reasons are concerns regarding worker exposure to ethylene oxide and the temperature limits of medical materials during high steam autoclaving. Radiation sterilization is a consequence of the high-energy electrons released from the interaction of gamma ray photons or high-energy electrons with materials. These secondary electrons in turn react with the DNA sequences in the microbiological burden in the device, permanently altering their chemical structure to render them innocuous.
Although the primary events of interaction with matter are different for gamma radiation and electrons, the major interaction in both cases is still Compton scattering, i.e., a shower of secondary electrons that initiates ionization events. These, in turn, activate numerous chemical reactions, many of which lead to oxidative degradation.
In addition to having an effect on harmful microbes, the high-energy electrons can also initiate ionization events in the material being sterilized, leading to the creation of peroxy and hydroperoxy free radicals in the presence of oxygen and the start of the degradation cascade. The results could be unacceptable color formation, pH shifts, and extractables. Furthermore, the degradation could also cause polypropylene to experience the well-publicized catastrophic failures during post-radiation shelf-life.
In earlier papers,(4) we documented significant oxidative stability reductions during gamma exposures of 20 and 40 kGys for flexible PVC, high-density polyethylene, and polypropylene. Because of the increasing usage of electron beam sterilization (beta irradiation), we are interested in how the radiation-induced degradation differs from one process to the other.
The oxidative degradation pathway for organic polymers typically follows an initiation, propagation, and termination sequence. The initiation, in this case the ionizing radiation, is followed by the propagation steps with atmospheric oxygen. It is noted that oxygen plays an autocatalytic role in the degradation cycle through the formation and subsequent decomposition of peroxide and hydroperoxide species. In addition, there are unimolecular and bimolecular termination reactions where free radicals recombine or disproportionate into neutral or inactive species (Eqs 3 to 5). At or near room temperature and under low dose rates, the active species concentrations are quite low; hence, the termination reactions are relatively unimportant. However, at higher temperatures where the polymer main chains are quite mobile or under high dose rate conditions, a high initiation rate causes the active species concentration to be quite high and the termination reactions to be very important.
R* + R* [right arrow] R-R (inactive) (3)
[RO.sub.2]* + R [right arrow] ROOR (inactive) (4)
[RO.sub.2]* [right arrow] Inactive species (5)
The high-density polyethylene film (Fig. 5) underwent a drastic reduction in OIT or antioxidant potency even at a near sterilizing dose of 30 kGy, from almost 900 min of induction time to about 20 min at 180[degrees]C. Subsequent additional doses resulted in a near linear decline in OIT on a logarithmic scale for the gamma radiation at a low dose rate of about 6 kGy per hour. A very significant reduction in OIT stability was also apparent in the electron beam samples at 30 kGy, although a slightly better retention of antioxidants can be clearly seen. From 60 to 100 kGy, a distinct upturn in OIT is plainly evident. This upturn is most likely caused by thermal and exposure time effects under high dose rates from the limited oxygen availability and accelerated free radical termination reactions favored at higher temperatures.
It is evident that processing can significantly degrade the material to a point where the molecular weight is reduced to a critical value leading to product failure. Material degradation can be prevented (or at least minimized) through an understanding of the material property performance and potential degradation paths. Careful selection of equipment and parameter optimization will ensure the retention of material properties at all times.
(1.) L. Woo, E. Chan, J. Palomo, and M.T.K. Ling, SPE ANTEC, 37, 2189 (1991).
(2.) J. Margolis, ed., Engineering Thermoplastics, Marcel Dekker, New York (1985).
(3.) A. Kinloch and R. Young, Fracture Behavior of Polymers, Elsevier, New York (1983).
(4.) S. Shang, M.T.K. Ling, S. Westphal, L. Woo, SPE ANTEC, 43, 2863 (1997).
PBT Rod Performance. Viscosity, Delta H, Annealing poise at J/g Temperature, 1 rad/sec at 6 Hr 240 [degrees]C PBT resin 33,000 N/A N/A Extruded rod, no annealing 24,000 41 N/A Extruded rod (run #30), annealed 2X 4800 54.7 300[degrees]F (148.9[degrees]C) Extruded rod (run #32), annealed 17,000 55.4 N/A Extruded rod (run #36), annealed 20,000 48.4 400[degrees]F Machined Drop Test [*] Clamp Behavior PBT resin N/A N/A Extruded rod, no annealing N/A N/A Extruded rod (run #30), annealed 2X brittle N/A Extruded rod (run #32), annealed O.K. N/A Extruded rod (run #36), annealed O.K. 50% break (*.)From 2 ft (60.94 cm), using a(204[degrees]C) 10-ft (304.8 cm) rod
SPE Books Related to This Topic:
The Effect of Sterilization Methods on Plastics and Elastomers (#1200)
Plastics Design Library, 1994, 478 pages, index, $241 (SPE member), $285 (nonmember).
Presented in graphical and tablular formats, the information in this publication includes trade names of materials, grades, suppliers, generic descriptions, qualitative assessments, fillers and additives used, and other details related to test specimens. Results of exposure data include changes in strength, elongation, impact, yellowness index, hardness, modulus, deflection temperature, vicat softening point, and electrical and optical properties.
Medical Plastics--Degradation Resistance and Failure Analysis (# 98325).
Edited by Robert C. Portnoy, 1998, 217 pages, $136 (SPE member), $160 (nonmember).
Providing an overview of the medical applications for plastics, along with specific performance requirements, this book focuses on the effects various degrading conditions, such as body liquids, harsh environments, and sterilization. The publication also discusses the failure of medical devices resulting from contamination, low temperature, UV light, migration of formulation components, mechanical stresses, and problems with design and fabrication.
|Printer friendly Cite/link Email Feedback|
|Comment:||How Processing Conditions Can Cause Degradation and Failure in Medical Devices.|
|Author:||Ling, Michael T.K.; Sandford, Craig; Sadik, Adel; Blom, Henk; Ding, Samuel Y.; Woo, Lecon|
|Date:||Jan 1, 2001|
|Previous Article:||Reduced-Emission Foam Surfactant.|
|Next Article:||Nanocomposites: The Importance of Processing.|