Effects of Electron Beam Sterilization on Polytetrafluoroethylene: Design of Experiments Study

Abstract

Polytetrafluoroethylene (PTFE) is a common polymer used in medical devices due to its exceptional properties (e.g., biocompatibility, inertness, chemical stability, low coefficient of friction). However, as a result of molecular weight reduction caused by the process of chain scission, it is known to be susceptible to radiation exposure and can rapidly lose strength and integrity. In this design of experiments study, the goal was to determine whether an operating window of conditions exist for electron beam (E-beam) radiation sterilization in which the degradation of PTFE is acceptably low. PTFE was tested for yield stress after exposure to radiation under different parameters (total dose [15–60 kGy], packaging atmosphere [air/nitrogen], and poststerilization accelerated aging [real-time equivalent of 1 and 3 years]). The results showed that total dose and packaging atmosphere were significant factors and indicated that the use of modified atmosphere packaging (vacuum sealing with nitrogen gas purge) can be a useful approach in increasing the stability of PTFE toward E-beam sterilization.

Gamma, electron beam (E-beam), and X-ray are forms of ionizing radiation that are widely used to sterilize medical devices, pharmaceutical products, and other materials. Radiation sterilization can achieve high levels of microbial reduction without requiring direct accessibility of all surfaces, and sterilization levels simply depend on the radiation dose received at a location. The inactivation of microorganisms by radiation and the use of radiation in sterilization processes are well understood and comprehensively documented in the literature.

However, the exposure of polymeric materials to ionizing radiation can alter their molecular structure and associated macroscopic properties.¹ These molecular changes are brought about through a complex set of reactions that generally culminate either in the net formation of cross-links or chain scission, resulting in an increase or decrease of molecular weight.² The chemical nature and morphology of the material determines which of these reactions are predominant.³

Polytetrafluoroethylene (PTFE) is a thermoplastic polymer that is commonly used in medical devices due to its biocompatibility, low friction coefficient, and high temperature and chemical resistance. These properties make it ideal for use in medical devices such as catheters, implants, and surgical instruments. However, PTFE is known to be susceptible to radiation, and its use in a device often means that radiation sterilization is not an option.

Factors Affecting Radiation's Influence on PTFE

The degradation of PTFE is especially pronounced in air.⁴ Irradiation in the absence of oxygen has much less of an effect on properties, as aerial oxygen promotes main-chain degradation by forming peroxyl radicals,⁵ which may prevent cross-linking.

A second factor that can influence the degradation of PTFE is the dose rate of a radiation incident. Diffusion of oxygen into a polymer may not be fast enough at high dose rates to replace the oxygen consumed, in which case radiation-induced oxidation will be more efficient at low than at high dose rates.⁶ As a result, the oxygen effect is much less pronounced in E-beam radiation (in kGy/s) than in gamma or X-ray irradiation (in kGy/h).

In addition, the effect of radiation on PTFE is expected to be dose dependent, as higher doses have been seen to cause greater degradation.⁷ Fourth and finally, the free radicals formed in PTFE due to radiation can be long lived.⁷

In this study, the effects of the above factors in the case of E-beam radiation of PTFE was investigated using a design of experiments (DOE) approach (Table 1). A radiation dose range of 15 to 60 kGy was evaluated in the study to capture the typical range of dose in which most medical devices can be sterilized, while meeting the twin requirements of microbial kill and material stability. Although radiation sterilization does not require breathable packaging (containing paper or a nonwoven material, such as synthetic flashspun high-density polyethylene fibers), radiation-sterilized devices often are still packaged in this manner due to various considerations (e.g., cost, printability, odor off-gassing poststerilization, high-altitude shipping compatibility).

Table 1.

Factors considered in the design of experiments.

In this study, breathable pouches were compared with foil pouches (which were vacuum sealed in a nitrogen atmosphere) to evaluate the effect of packaging environment (air versus oxygen free). All irradiated samples were also aged for the real-time equivalent (RTE) of one and three years to evaluate the possible effects of postirradiation aging. All possible combinations of the factors were tested. The center point for radiation dose was determined to be 37.5 kGy, and four repeats for the center point were conducted (using the four possible combinations of the other factors: packaging environment and postexposure accelerated aging; Tables 1 and 2) to evaluate process stability and inherent variability.

Table 2.

Yield stress for samples as a function of different factors. Abbreviation used: NA, not applicable.

Materials and Methods

Samples of nonirradiated PTFE heat-shrink tubing (internal diameter 0.04 in, wall thickness 0.007 in, length 6 in) were prepared and put into breathable pouches (porous to air) or foil pouches that were vacuum sealed (with nitrogen gas purge) to represent the air and nitrogen packaging atmosphere variables, respectively. Of note, PTFE heat-shrink tubing is widely used in the medical device industry for various applications (e.g., guidewire jacketing, coating surgical instruments, dielectric insulation, waterproofing, abrasion protection, dielectric insulation, reflow bonding).

The sealed pouches were placed in shelf cartons, arranged in a shipper box, and sent for E-beam sterilization (at SteriTek in Fremont, CA) at different targeted doses (15, 35, and 60 kGy). After irradiation, samples were aged in a chamber under accelerated aging conditions (55°C and 50% relative humidity) for a RTE of one and three years, per the Arrhenius equation,⁸ assuming a Q₁₀ factor of 2.0.

Samples were tested on a tension fixture using a dynamic mechanical analyzer (RSA-G2 DMA; TA Instruments, Newcastle, DE). Approximately 10 mm sections were cut from the PTFE samples for testing. A stress constant (linear) of 1,000 kPa/N and a strain constant (linear) of 200 m^–1 were used. (These are the default instrument values and cannot be changed.) A preload of 10 gf was applied to each sample before the test was initiated, with the test conducted at a constant linear rate of 0.0417 mm/s (~25% of the sample length/min). Yield stress (in MPa) was obtained from the tensile curves generated and used as an output/response of interest for the DOE. Design-Expert (version 10.0; Stat-Ease, Inc., Minneapolis, MN) was used to generate and analyze the DOE. A sufficient number (n > 5) of viable samples for some of the conditions were not available and were excluded from the analysis.

Results and Discussion

Table 2 summarizes the results for averaged yield stress values (n > 5) for the different samples. A clear dose dependence in yield stress values can be seen for samples that were packaged in breathable pouches (exposed to air) irrespective of poststerilization accelerated aging. Of note, a decrease in yield stress would indicate the deleterious effect of radiation chain scission on PTFE samples. However, samples that are packaged in foil pouches, along with a nitrogen packaging atmosphere, are seen to better retain their yield stress values with increasing radiation dose.

A Box-Cox plot (Figure 1) was used to probe the nature of the response data (yield stress) for the DOE⁹ and suggested a logarithmic transformation for the response data. As a result, the logarithm of the response data was used as input for the DOE analysis. Table 3 provides a summary of an analysis of variance of the transformed data and lists the elements/factors in the DOE that were found to be significant (P < 0.05). The radiation dose, packaging environment, and their mutual interaction (represented by "AB" in Table 2) were found to be highly significant (P < 0.0001), which is consistent with the data seen in Table 2. The A² factor, which represents the quadratic nature of the yield stress versus dose (Figure 2), was also found to be significant.

Figure 1.

Box-Cox plot for yield stress data. Abbreviations used: CI, confidence interval; SS, sum of squares.

Table 3.

Analysis of variance for different factors in the design of experiments. Abbreviation used: df, degrees of freedom.

Figure 2.

Predicted yield stress (averaged; n > 5) of polytetrafluoroethylene (PTFE) samples as a function of radiation dose and packaging environment under the conditions of 1 year (A) and 3 years (B) real-time equivalent accelerated aging. The blue horizontal reference line indicates the baseline value for nonradiated, nonaged samples. Green and red dotted lines indicate the 95% confidence intervals for predicted yield stress values as a function of dose.

Figure 2 shows that the predicted yield stress for samples that were packaged in a nitrogen atmosphere in foil pouches remained close to the baseline value for the nonirradiated, nonaged samples up to 46 kGy of radiation dose and decreased by approximately 10% at 60 kGy of radiation dose. However, for samples exposed to air, a decrease in yield stress of more than 35% was seen at 15 kGy exposure and a decrease of more than 90% in yield stress was seen at 60 kGy exposure. A comparison of Figure 2A and B confirms visually that accelerated aging did not affect the yield stress values significantly, which was consistent with the high P value (>0.05) for the factor in Table 3.

These results demonstrated that the packaging of medical devices containing PTFE in an oxygen-free environment can have a considerable effect on the performance and ability of devices to maintain desired material properties when a radiation-based modality is considered for sterilization.

Conclusion

In conclusion, E-beam has a significant dose-dependent effect on the properties of PTFE, as shown by a 35% to 90% reduction of yield stress in the aerial environment described here. However, the use of an oxygen-free environment in packaging can significantly reduce the degradation of PTFE, which retains its yield stress up to 46 kGy of radiation dose and undergoes only a 10% reduction in yield stress at 60 kGy of radiation dose.

With the capacity and environmental concerns around the use of ethylene oxide sterilization and the limited availability of other viable alternative modalities, novel approaches are needed to increase the material compatibility of medical devices with radiation sterilization. The use of modified-atmosphere packaging, coupled with dose optimization, could increase the range of medical devices that can be sterilized using radiation.