The Effects of Radiation Sterilization on the Microbial Barrier Properties of Tyvek

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI October 1999 Column

A study showed that, for Tyvek, the microbial barrier properties follow filtration theory for the variables of basis weight, porosity, and charge.

Although the porous materials used in the packaging of terminally sterilized medical devices must allow air and moisture to flow into the package during certain sterilization processes, their primary requirement is to block the passage of microorganisms that would compromise device sterility. A number of test methods have been used to measure such barrier properties, with varying degrees of success. Some of these include the ASTM F 1608 95 standard test method for microbial ranking of porous packaging materials (also known as the exposure chamber method), the methylene blue aerosol test, and various tests using microorganisms bound to particles of talc or dust. For the study reported on in this article, the method developed by Alan Tallentire and his colleagues at Air Dispersions Ltd. (Manchester, UK) was chosen. The study was conducted to measure the effects of a range of sterilization methods on the microbial barrier properties of Tyvek. The results of radiation sterilization tests are presented here; the results of tests using ethylene oxide, chlorine dioxide, steam, Plazlyte, and Sterrad sterilization will be discussed in a future issue.

METHODS AND MATERIALS

A complete description of the method used to measure microbial barrier performance was published in the May 1996 issue of MD&DI.¹ The various steps involved can be summarized as follows. First, endospores of Bacillus subtilis var. niger are aerosolized into a rotating chamber that allows the spores to remain as discreet particles in a dry state and maintains a uniform dispersion over time. Then a sample of test material is placed in an apparatus located downstream of the rotating chamber and upstream from a gelatin filter. Air containing a known concentration of spores flows through the sample and filter at defined rates. Airflow is controlled using critical flow orifices, and the spores that are trapped on the filter are enumerated using standard microbiological techniques. The microbial barrier properties of the test material, which are described as percent penetration, are determined by comparing the number of spores that pass through the sample with the known challenge concentration. Because the percent penetration will differ at various flow rates, when multiple flow rates are used, a curve is generated that defines the rate at which the maximum percent penetration occurs. This rate is defined as P_max.

Table I. Styles of Tyvek evaluated in the study.

All test samples used in this study were 12.7 cm in diameter, and airflow rates ranged from 2.17 x 10^–4 to 1.32 x 10^–1 dm³ min^–1 cm^–2 (L/min/cm²). The rate of airflow through the sample was measured at a fixed pressure difference of 20-cm water pressure (which allowed air permeance measurement without destruction of the sample). The measurement was then normalized to Bendtsen units (cm³ min^–1 at 10-cm water pressure for a test sample area of 10 cm²) using a standard curve. Sample basis weights were determined to three decimal places using an analytical balance. The types of materials evaluated in the tests are listed in Table I.

TESTS, RESULTS, AND DISCUSSION

Exposure to radiation is known to reduce or eliminate charge in a wide variety of materials. In the first part of this study, the effect of the inherent charge of the high-density polyethylene base polymer on the microbial barrier properties of the test materials before and after irradiation was determined. In these tests, five circular samples were cut at random from a sheet of Tyvek 1073B. Each sample was measured for basis weight, followed by measurements for spore penetration at five flow rates and Bendtsen permeance. These results served as the control (0-kGy exposure) data.

Figure 1. Microbial barrier profiles of Tyvek 1073B: (a) unexposed to gamma irradiation, (b) after a 2-kGy exposure, (c) after a 4-kGy exposure, (d) after an 8-kGy exposure, (e) after a 16-kGy exposure, and (f) after a 32-kGy exposure.

Each sample was then placed in an open beaker, and irradiated with gamma rays from a Mark IV Hotspot irradiator. A ferrous sulfate dosimeter was used to measure the average dose rate in the beaker. Following this procedure, the sample was again measured for basis weight, followed by measurements for spore penetration at five flow rates and Bendtsen permeance. This irradiation and measurement process was repeated sequentially to provide data for samples that had received total doses of 2, 4, 8, 16, and 32 kGy, respectively. Figures 1a–f demonstrate that upon irradiation the material's microbial barrier properties shifted to a lower P_max and higher percent penetration. However, this shift did not appear to be dose dependent.

Figure 2. Microbial barrier profile of Tyvek 1073B: (a) before and after 8-kGy gamma irradiation and (b) before and after 8-kGy E-beam irradiation.

To determine if this effect is specific to gamma irradiation, electron-beam (E-beam) exposure testing was conducted using the same methods except that noncontrol samples received only one level of radiation exposure—8 kGy. In addition, 10 samples were cut from the sheet of Tyvek 1073B and placed in a closed petri dish as a single stack. E-beam irradiation was carried out with a 10-MeV LINAC operating at an average of 1 mA and with a 60-cm scan width on a single-pass conveyor. The dose delivered to the samples was monitored using a water calorimeter. Figures 2a and 2b demonstrate that following E-beam processing, a shift in microbial barrier properties occurred that was identical to that resulting from gamma irradiation.

Figure 3. Microbial barrier profile of Tyvek 1073B manufactured in Luxembourg before and after 8-kGy gamma irradiation.

To determine if this general radiation effect is dependent upon the polymer source, manufacturing process, or basis weight of the material, samples of Tyvek 1073B manufactured in Luxembourg and Tyvek 1059B, which has a lower basis weight than 1073B, were tested as described at an 8-kGy gamma radiation exposure. As seen in Figure 3, the material manufactured in Luxembourg, which was made with polymer from a different source than the 1073B evaluated in the earlier tests, and processed with 64 rather than 32 spinning positions, was affected similarly. The radiation effect was also evident when samples of 1059B were tested (Figure 4).

Figure 4. Microbial barrier profile of Tyvek 1059B before and after 8-kGy gamma irradiation.

Previous research had indicated that minute changes in a polymer's charge could not be measured accurately. Therefore, to determine if the radiation effect was, in fact, due to charge, we investigated the hypothesis that by increasing the charge of the base polymer, the microbial barrier properties of the material would be improved and that, after subsequent irradiation, those properties would demonstrate the same radiation shift as untreated materials. To perform this experiment, Tyvek 1073B was subjected to on-line corona treatment during manufacturing. This surface treatment is not part of the normal manufacturing process for this material because there is a risk of pinhole formation resulting from electrical arcing across the electrodes. However, it is used to produce Tyvek 1073D, a commercial product for the graphics industry that has the same basis weight as 1073B but a different porosity specification.

Figure 5. Microbial barrier profiles of experimental materials before and after 8-kGy irradiation: (a) Tyvek 1073C 6/0 and (b) Tyvek 1073C 6/6.

Two experimental corona-treated samples were produced. The first (1073C 6/0) was cut from sheet that had been treated on only one side, while the second (1073C 6/6) was from sheet that had been treated on both sides. These samples were subjected to the test protocol and gamma irradiation exposure of 8 kGy. The results are presented in Figures 5a and 5b. Both unirradiated corona samples were tested for microbial barrier properties at all five rates. However, increasing the charge of the polymer had improved the barrier properties to such an extent that no microorganisms were detected downstream of the unirradiated test samples at the lower flow rates. Therefore, no data points are shown for some flow rates.

Two conclusions can be drawn from these findings. First, the microbial barrier properties of Tyvek are improved by corona treatment, and such improvement appears to be dose dependent since the sample treated on both sides retained more spores than the sample treated on only one side. In addition, downstream spores were detected at only the two highest flow rates for the sample with two-sided treatment versus at three flow rates for the other sample. Second, a postirradiation shift occurred for both experimental samples and was of the same general magnitude.

Figure 6. Microbial barrier profile of Tyvek 1073D before and after 8-kGy gamma irradiation.

Another experiment was designed to test the hypothesis that treating a material with an antistat should have the same effect as irradiation if, in fact, the polymer charge was neutralized. Tyvek 1073D was used to test this hypothesis. This material has the same basis weight as 1073B, but it has a different porosity and undergoes corona and antistat treatments during manufacturing. The samples were subjected to the test protocol described above using 8-kGy gamma radiation. The results (Figure 6) indicated that radiation has no effect on the microbial barrier properties of 1073D that can be attributed to the polymer charge reduction caused by the antistat treatment. It is important to note that data for 1073B and 1073D should not be compared because the materials have different porosity specifications.

Figure 7. Microbial barrier profile of a 55-lb medical-grade paper before and after 8-kGy gamma irradiation.

A final experiment was conducted to determine whether a similar radiation shift would be demonstrated with paper substrates used in medical packaging. Samples of an uncoated 55-lb medical-grade paper were tested using 8-kGy gamma radiation. The control samples exhibited barrier properties inferior to those of any of the Tyvek samples tested prior to irradiation, and the irradiated samples showed no evidence of the irradiation effect (Figure 7). A 70-lb medical-grade paper was also tested in the unirradiated state, and it, too, demonstrated barrier properties inferior to those of Tyvek.

It has been proposed that porous packaging materials function as depth filters to maintain the sterility of a medical device. A good example of that type of filter is the HEPA filter commonly used in cleanroom air systems. The efficiency of a HEPA filter is a function of several variables including the nature of the substrate, distribution of fiber diameters, fiber charge, filter depth (a function of basis weight and porosity), pressure differential (flow rate) across the filter medium, and size of the particles used in the test. Because the flow rate (face velocity) for such a filter is specified by the filter manufacturer, there is no question about the most appropriate rate at which to test a sample. However, for medical packaging, distribution and handling are the primary influences on the pressure differentials across a sterilized package, so the establishment of definitive test parameters remains a challenge.

This study provides clear evidence that, for Tyvek, the microbial barrier properties follow filtration theory for the variables of basis weight, porosity, and charge. To adequately test the remaining variables, it would be necessary to develop experimental samples that have differing fiber-diameter distributions and are made from polymers from different sources. Some studies of this type are under way, and we hope to report on their results in the future.

CONCLUSION

The results of the experiments described above can be summarized as follows:

ACKNOWLEDGMENT

The authors wish to thank Alan Tallentire and Colin Sinclair for their thoughtful insights into this study. We would also like to thank the laboratory staff of Air Dispersions Ltd., which performed the testing described in this paper, and Peggy Rook for her assistance in generating the test samples.

REFERENCE

1. Alan Tallentire, "A Discriminating Method for Measuring the Microbial Barrier Performance of Medical Packaging Papers," Medical Device & Diagnostic Industry 18, no. 5 (1996): 228–241.

All of the authors are affiliated with DuPont Nonwovens, serving at its facilities in Wilmington, DE; Richmond, VA; and Luxembourg. Michael Scholla, PhD (Wilmington), and Claude Michels (Luxembourg) are business managers for the medical packaging segment of the Tyvek enterprise within DuPont Nonwovens, with responsibilities for the Americas and Europe, respectively. Earl Hackett (Wilmington) and Wazir Nobbee and Stasys Rudys (Richmond) make up the manufacturer's medical packaging technical team.

Back to the Table of Contents

Copyright ©1999 Medical Device & Diagnostic Industry