Effect of gas composition on spore mortality and etching during low-pressure plasma sterilization

Effect of gas composition on spore mortality and etchingduring low-pressure plasma sterilization

S. Lerouge,1 M. R. Wertheimer,2 R. Marchand,3 M. Tabrizian,1 L’H. Yahia1

1Research Group on Biomechanics and Biomaterials, Biomedical Engineering Institute, Ecole Polytechnique, StationCentre-Ville, P.O. Box 6079, Montreal, Quebec H3C 3A7, Canada2Department of Engineering Physics and Materials Engineering, Ecole Polytechnique, Station Centre-Ville, P.O. Box6079, Montreal, Quebec H3C 3A7, Canada3Montreal Heart Institute, 5000 Belanger, Montreal, Quebec H1T 1C8, Canada

Received 1 June 1999; revised 26 October 1999; accepted 1 December 1999

Abstract: The aim of this work was to investigate possiblemechanisms of sterilization by low-temperature gas plasma:spore destruction by plasma is compared with etching ofsynthetic polymers. Bacillus subtilis spores were inoculatedat the bottom of glass vials and subjected to different plasmagas compositions (O2, O2/Ar, O2/H2, CO2, and O2/CF4), allknown to etch polymers. O2/CF4 plasma exhibited muchhigher efficacy than all other gases or gas mixtures tested,with a more than 5 log decrease in 7.5 min, compared witha 2 log decrease with pure oxygen. Examination by scanningelectron microscopy showed that spores were significantlyetched after 30 min of plasma exposure, but not completely.

We speculate about their etch resistance compared with thatof synthetic polymers on the basis of their morphology andcomplex coating structure. In contrast to so-called in-houseplasma, sterilization by Sterradt tended to increase the ob-served spores’ size; chemical modification (oxidation),rather than etching, is believed to be the sterilization mecha-nism of Sterradt. © 2000 John Wiley & Sons, Inc. J BiomedMater Res, 51, 128–135, 2000.

Key words: sterilization; low-pressure plasma; mechanismof spore destruction; etching; O2/CF4

INTRODUCTION

Gas plasma technology is increasingly being used toetch and modify polymeric surfaces1; a less commonapplication is its use as an alternative low-temperaturesterilization method for biomedical devices. The in-creasing development of complex, delicate polymericdevices for medical applications, and the presence ofhazardous residues and by-products left by chemicalsterilization—for example, when using ethylene oxide(EO)2—have underlined the critical need for safer, al-ternative low-temperature sterilization methods. Thishas greatly raised interest in low-pressure gas plasmasterilization in recent years.

Very succinctly, cold plasma is a partially ionized,low-pressure gas comprising of ions, electrons, and

ultraviolet (UV) photons, as well as reactive neutralspecies (radicals and excited atoms and molecules),with sufficient energy to break covalent bonds andinitiate a variety of chemical reaction pathways.Plasma has been shown capable of destroying micro-organisms.3–10 It is a promising technology in that itacts rapidly, it does not leave toxic residuals on pro-cessed parts or in the exhaust gas, and the temperatureof a substrate usually does not exceed 60°C. Two ster-ilizers using plasma technology have been commer-cialized: namely, Sterradt (Advanced SterilizationProducts, Johnson and Johnson, Arlington, TX) andthe Plazlytet Sterilization System (AbTox, Mundelein,IL). However, little is known or has been published sofar about their precise mode of action. Both methodscombine the use of an oxidative chemical phase intheir sterilization cycles, in addition to plasma. Theplasma sterilization principle of Sterrad-100St consistsof diffusing vaporized H2O2 in the treatment chamberbefore applying 300 W of radiofrequency (RF) powerat a pressure of 0.5 Torr to create the plasma. Plazlytetuses peracetic acid vapor exposure, which is alter-nated with a downstream plasma treatment by micro-wave (MW) excitation of a low-pressure gas mixture

Correspondence to: L’H. Yahia; e-mail: [email protected]

Contract grant sponsor: Natural Sciences and EngineeringResearch Council of Canada

No benefit of any kind will be received either directly orindirectly by the authors.

© 2000 John Wiley & Sons, Inc.

composed of O2, H2, and inert gas. Owing to the twodistinct treatment phases (chemical and plasma),many questions still remain—for example, regardingthe inherent bactericidal efficacy of either phase,whether they act in synergy, and the efficacy ofplasma through sterilization wrapping. Recently,Krebs et al.11 reported that the measured efficacy ofthe Sterrad-100t process is above all due to hydrogenperoxide, the chemical agent that is left to diffuse for50 min before its destruction by the action of plasmaexposure.

More generally, the efficacy and mechanisms ofplasma processes upon microorganisms are stilllargely unknown. Most of the available data comefrom the patent literature, and results are often con-tradictory.3,4,8,10,12 Among specific physicochemicalmechanisms of plasma which can possibly kill micro-organisms, the following have been suggested: physi-cal sputtering of the outer cell walls, chemical degra-dation by reactive species, and/or by UV light fromthe plasma.8,10,12 As also pointed out by Pelletier,13

microorganisms are in some ways similar to syntheticpolymers—namely, macromolecules composed of theelements C, H, N, and O; it is therefore interesting todraw parallels with the abundant literature on plasmatreatment and etching of polymers. Many gases andgas mixtures can etch polymers in low-pressureplasma, the common feature being that they convertthem into volatile reaction products, which are thenremoved by the vacuum pump. Many of the samegases have also been shown to have some efficacy inspore destruction.3,5–7,9 Because polymer etch rates ex-ceeding 1 mm/min can be achieved with plasma,14–16

it is interesting to examine the possibility that simpleetching (volatilization) may be one of the main mecha-nisms of spore destruction by plasma.

The objective of this work was to study the effect ofgas plasma on bacterial spores. More specifically, ourfirst goal was to compare the efficacy of differentplasma gas compositions, all known to etch polymers.To further help us elucidate possible kill mechanisms,we looked for the visible effect on spores by scanningelectron microscopy (SEM) and compared the resultsof treatments in laboratory plasma with those of Ster-radt, pure EO, and steam sterilizations.

MATERIALS AND METHODS

Sample preparation

Bacillus subtilis (Bs) spores were chosen because they areknown to have high resistance to plasma, and because theyare used as Sterradt biological indicators.5,10,17 Bs spores insuspension in a 40% ethanol solution (ATCC 9372, lot

973087; Spordext, AMSCO, Erie, PA) were used in thisstudy. The D-values (time necessary to decrease the numberof spores by 1 log) of these spores in dry heat or in pureethylene oxide were 1.3 and 3.8 min, respectively, accordingto the supplier. One hundred microliters of the suspensionwas aseptically spread on the flat bottom of special borosil-icate glass vials (F = 25 mm, h = 18 mm) in the form of amonolayer, and left to dry. Preliminary studies showed thatthe initial number of spores and their dispersion and con-centration (that is, their distribution) influenced the results.Therefore, we kept the initial number of spores (107) and thesurface area (200 mm2) constant to ascertain a monolayer ofspores for each sample. For reasons explained in the Discus-sion, we also carried out experiments with polystyrene (PS)microspheres (1 mm in diameter) in the place of the Bsspores.

Plasma exposure

Plasma sterilization treatments were carried out in a large-volume microwave plasma (LMP™) reactor, described else-where14 and shown schematically in Figure 1. Microwavepower (MW 2.45 GHz) was applied from a strapped-barslow-wave structure through a rectangular fused silica win-dow into the stainless-steel reactor chamber. The groundedstainless-steel sample holder was maintained parallel to thewindow so that the open ends of the vials faced the glowzone. For each experiment, three vials were placed into theplasma reaction chamber.

Oxygen-containing gases or gas mixtures used for thisinvestigation were the following: (a) pure O2; (b) O2/H2

(20/80%); (c) O2/Ar (50/50%); (d) O2/Ar/H2 (33% each); (e)pure CO2; and (f) O2/CF4 (various mixture ratios). Thesewere chosen on the basis of earlier literature on sterilizationand on polymer etching, as discussed below. Oxygenplasma is used widely for the removal of organic materials—for example, for the removal of photoresists after litho-graphic operations in microelectronics1—and the etch rate isstrongly correlated to the concentration of oxygen atoms[O], produced by the dissociation of O2.18 Addition of inertgas such as argon (Ar) is known to increase the yield of

Figure 1. Schematic representation of the in-house micro-wave plasma.

129MECHANISM OF PLASMA STERILIZATION

atomic oxygen in O2 plasma,1,19 whereas the addition of H2

increases the output of vacuum ultraviolet (VUV) radiationby the plasma.20 Finally, CF4 can play an important catalyticrole in the etch process and greatly increase the etch rate ofpolymers.1,14,15 CO2 was also chosen because it has demon-strated higher bactericidal efficiency than pure O2.10 Each ofthese gases was introduced into the chamber at an operatingpressure of 80 mTorr, and plasma was excited with 200 W ofMW power. Spores were exposed to the plasma for 2.5, 7.5,and 15 min, the treatment time (t). To minimize possibletemperature increase (that is, to separate thermal andplasma effects), the plasma was applied in the form of 30-spulses followed by 30-s pauses. Among the many experi-mental plasma parameters, we chose to study the influenceof flow rate (F), as well as the distance between the spore-bearing substrate and the microwave window (d), whereboth these factors are known to influence the types and con-centrations of plasma species reaching the samples. How-ever, for most experiments, d was fixed at 6 cm, smallenough for the plasma glow to penetrate into the volume ofthe vials, and F was maintained at 50 standard cm3/min(sccm).

After treatment, the spores were recovered aseptically inbrain heart infusion medium, serially diluted, and spread onblood agar. The number of colonies regenerated wascounted after 24 h of incubation at 37°C. The log reduction ofspore count, also called mortality (M), was calculated byaveraging log (No/N) for given treatment conditions, whereNo is the initial number of viable spores (calculated fromthree control vials) and N is the number of viable sporesafter exposure to plasma. From these data, M was plottedversus t.

SEM analysis

Spores were inoculated on glass and on single crystal sili-con substrates, and subjected to 15 min of plasma treat-ments, as previously described. A thin, conductive film ofgold-palladium was then deposited onto the surface, afterwhich the treated and untreated spore samples were ob-served by SEM (Jeol; JSM 840). The effects of the in-houseplasmas with different gas mixtures were compared withthose induced by commercial sterilization techniques suchas Sterrad-100S®, pure ethylene oxide (SteriVac®; 3M), andsteam (AMSCO, 3M; 20 min at 121°C). Samples for commer-cial sterilization were added to normal sterilization loads inhospital facilities before their observation by SEM.

RESULTS

Sporicidal efficacy of plasma treatments

The relative efficacies of the six different plasmatreatments on the Bs spores are compared in Figure 2.All treatments induced a spore mortality of 2 log or

more in 15 min, with pure O2 plasma being the leasteffective treatment. Little difference was noted be-tween O2, O2/H2, O2/Ar, O2/Ar/H2, and CO2 plas-mas, whereas O2/CF4 plasma exhibited a much higherefficacy: namely, a 5 log decrease within t = 7.5 min,compared with the 1–2 log decreases with the othergases or gas mixtures, for the same value of t. It istempting to attribute this to a higher etch rate of O2/CF4 plasma; indeed, the admixture of a fluorine com-pound to oxygen is known to greatly enhance the etchrate of most polymers.14 The polymer etch rate variesas a function of the CF4 concentration [CF4] in theO2/CF4 mixture. Therefore, we specifically investi-gated the effect of [CF4] on spore mortality, M. Figure3 shows that the variation of M with [CF4] followed aremarkably similar trend to that of the etch rate, R, ofa particular polymer (Kapton® polyimide), as foundin the same reactor by Lamontagne et al.14 Both curvespresented clear maxima near [CF4] = 12%. We there-fore hypothesize that the sporicidal rate in O2/CF4plasma is directly related to the etch rate of the naturalmacromolecules which make up the microorganisms.The influence on M of gas flow rate (F) and the dis-tance between the spore-bearing substrate and theMW window (d) are presented in Figures 4 and 5,respectively. As expected intuitively, M decreasedwith increasing d. In contrast, M increased with in-creasing F, reaching a value of more than 6 log in t =5 min. Because the behavior of spore mortality withtime is not always linear, it is not appropriate to con-vert these results into real D-values; however, theequivalent D-value for this latter plasma treatmentwas about 0.8 min, compared with 1.3 and 3.8 min fordry heat and pure EO, respectively.

Figure 2. Spore mortality, M, of microwave plasma as afunction of gas composition (p = 80 mTorr; P = 200 W; F = 50sccm; d = 6 cm) l = O2/CF4 (85/15%); j = O2/Ar (50/50%);m = O2/H2 (20/80%); h = O2/H2/Ar (33% each); n = CO2;d = O2.

130 LEROUGE ET AL.

SEM

Scanning electron microscopic observation con-firmed the etching effect of spores by O2-containingplasmas, particularly by O2/CF4 plasma. Control Bsspores were ellipsoidal, with average dimensions of1.2 mm × 0.6 mm [Fig. 6(a)]. Whatever the gas mixture,

spores exposed to plasma for 15 min were smallerthan the controls, indicating some etching [Fig. 6(b)],whereas spores exposed to O2/CF4 plasma were moststrongly etched, as seen in Figure 6(c) by their muchsmaller structures and microscopic debris. Neverthe-less, the spores exhibited a higher resistance to plasmaetching than expected: Even after t = 90 min, sporeswere not completely removed and solid residuescould still be observed.

The effects of the in-house plasma treatments werecompared with those induced on Bs spores by thecommercial sterilization methods (Sterrad-100St,steam, and pure EO); the micrographs shown in Fig-ure 6(d–f) present their characteristic effects. Thevariation of the number of spores seen on these mi-crographs is mainly due to nonuniformity of the sporelayer on the substrates before treatment. Clearly, thereis no size reduction visible in any of these micro-graphs. Therefore, spore mortality here cannot be at-tributed to etching or erosion, as discussed subse-quently. Absence of etching during the Sterrad® ster-ilization process was confirmed by SEM examinationof small latex microspheres (1 mm) subjected to Ster-rad-100St.

DISCUSSION

Efficacy and mechanism of O2-containing plasmas

We compared the sporicidal efficacy of variousplasma treatments, all known to etch polymers. O2/CF4 plasma exhibited much higher efficiency in killingBs spores than all other plasma treatments tested, par-ticularly the mixture with [CF4] = 12%. To the best of

Figure 3. Spore mortality and polymer etch rate as a func-tion of [CF4] in an O2/CF4 microwave plasma. j = sporemortality, M, induced by t = 5 min of plasma treatment (p =80 mTorr; P = 200 W; 50 sccm; d = 6 cm); l = etch rate, R(mm/min) of Kapton® polyimide film, in the same reactor(after Lamontagne et al.14).

Figure 4. Spore mortality (M) of O2/CF4 plasma as a func-tion of the gas flow rate F (p = 80 mTorr; P = 200 W; d = 6 cm;t = 5 min; [CF4] = 8%).

Figure 5. Spore mortality (M) of O2/CF4 plasma as a func-tion of the distance between the bottom of vials and themicrowave applicator, d (p = 80 mTorr; P = 200 W; F = 50sccm; t = 5 min; [CF4] = 8%).

131MECHANISM OF PLASMA STERILIZATION

our knowledge, we have demonstrated for the firsttime the very high efficacy of this mixture for steril-ization. The sporicidal rate we observed for O2/CF4,6.3 log in 5 min (which corresponds to more than99.9999% destruction), was higher than that of allother gases or gas mixtures we tested, and higher thanvalues reported by other authors.5,7,8,10 Lin5 foundN2O to be more efficient than O2 and CO2, but heobtained <5 log reduction after 30 min. H2O2 plasma

was found even more efficient,5 but this is a subject ofcontroversy, as there is reason to believe that this ef-ficacy was due to the sporicidal properties of H2O2vapor rather than to the plasma itself.10,11 Indeed,when H2O2 was not left to diffuse before creatingplasma, it was less effective in killing spores thanCO2.11 The efficacy of the low-power RF argon plasmaof Baier et al. was limited to a 3 log decrease in 9 min.7

Chau et al.8 found that 105 Bacillus stearothermophilus

Figure 6. Scanning electron micrographs of Bacillus subtilis spores: (a) untreated control; (b) spores exposed to t = 15 min ofpure O2 plasma; (c) spores exposed to t = 15 min of O2/CF4 plasma (p = 80 mTorr; P = 200 W; F = 70 sccm; d = 6 cm [CF4]= 15 %); (d) spores treated with Sterrad-100S®; (e) spores treated by steam autoclaving (20 min at 121°C); and (f) sporestreated by pure EO.

failed to grow only after 20 min of N2O electron cy-clotron resonance plasma treatment. Hury et al.10 ob-served that CO2 plasma exhibited a higher destructionefficiency than its pure oxygen counterpart, but spo-ricidal activity was limited to 6 log in 30 min.

Our study demonstrates that gas mixture and flowrate greatly influence the rate of spore destruction.However, other parameters can also explain the effi-cacy of our system, especially the choice of MWpower. We conducted similar plasma sterilization ex-periments in the same apparatus, in which the MWpower source for plasma excitation was replaced by13.56 MHz RF power; those results will be reportedelsewhere. As in the case of polymer etching,15,16 theefficacy of RF plasma was found to be lower than itsMW counterpart under otherwise identical conditions.The reasons for the observed differences between MWand RF plasmas have been the object of in-depth in-vestigation in these laboratories.22 They are basicallyattributed to different electron energy distributionswhich, of course, strongly affect the concentrations ofcharged and neutral active species. The relatively highpower employed (similar to that of commercial Ster-rad® and Plazlyte systems, but higher than in someother work7,8) and the optimum operating pressure, asdetermined by results on polymer etching, also helpexplain our results.

Nevertheless, there are several reasons why directcomparisons between published studies are generallynot straightforward: (a) the behavior of spores underplasma treatment is not always linear8,10; (b) sporeresistance depends not only on the spore type, but alsoon the methods of preparation and regeneration, onthe material and geometry of the spore carriers, andon their concentration and distribution on the carriers;and (c) plasma exposure parameters—for example,the distance between samples and discharge source—also influence results, as we clearly observed in thepresent work.5,10,11 Unfortunately, these parametersare not always clearly specified in publications andpatents. It would be helpful if they were specified and,ideally, standardized for future studies.

As already stated, our results strongly suggest thatthe high sporicidal efficacy of O2/CF4 plasma is linkedwith its high etch rate of organic solids.1,14,15 There-fore, it is interesting to relate our results to the litera-ture on polymer etching. To enhance polymer etching,plasma parameters can be selected to increase the dis-sociation of O2 to atomic oxygen, decrease recombina-tive loss of O, and increase the flux of O from theplasma to the sample.1 This can explain, for example,why we found spore mortality, M, to increase with thegas flow F and to decrease with the distance d. Al-though argon metastables are known to play an im-portant role in the overall excitation mechanism of O2plasmas, O2/Ar plasma did not demonstrate muchgreater efficiency than pure O2 plasma. Neither did

the O2/H2 plasma, although it is known for its in-creased output of highly energetic VUV radiationcompared with pure O2 plasma. This suggests that thesporicidal action of plasma is not greatly affected byVUV radiation. Further investigation of this contro-versial issue is currently in progress.

In contrast to Ar and H2, the addition of CF4 greatlyenhanced spore mortality because fluorine-containingadditives such as CF4 enhance etching by two differ-ent pathways (Fig. 7). First, they are known to increasethe oxygen atom concentration; second, and more im-portant, fluorine atoms produced via the dissociationof CF4 play a catalytic role in the etch process by weak-ening the polymer. F atoms can abstract hydrogenfrom organic solids through highly exothermic reac-tion, after which the resulting radical sites are readilyattacked by O atoms, or possibly even by molecularoxygen.1 Backbone bonds near the radical sites arethereby significantly weakened, and this reaction se-quence leads to the observed volatile reaction prod-ucts (HF, CO, CO2, COF2, . . . ). The catalytic effect ofCF4 can explain the higher efficacy of O2/CF4 plasmacompared with the other gas compositions studiedhere and by other authors.

Although SEM observations of spores generally donot allow one to draw conclusions about the mecha-nism of a given sterilization processes, they clearlyshow that etching contributes to the action of O2-containing plasmas, particularly of O2/CF4 plasma.However, the Bs spores exhibited a surprisinglyhigher resistance to etching than expected; even after90 min of O2/CF4 plasma, they were not totally re-moved. Yet, the initial spore layer thickness in thevials was only a monolayer, <1 mm. Under the sameetch conditions, a 1-mm-thick film of synthetic poly-mer would have been completely removed in at mosta few minutes in this same apparatus.14,15 There may

Figure 7. Schematic representation of etching in a low-pressure oxygen plasma: enhancement by addition of a fluo-rine-containing gas.

133MECHANISM OF PLASMA STERILIZATION

exist several possible explanations for this unexpect-edly high resistance. The first is related to the self-repairing ability of spores, and to their particularstructure and composition. Unlike inanimate syntheticpolymers, microorganisms are capable of self-repair,so that damage is not always irreversible. Preliminarymicrobiological results showed that the number andsize of spore colonies regenerated on gels depend ontime after plasma treatment, which suggests that somefraction of the viable spores damaged by the plasmawas capable of recovery. Moreover, the basic structureof spores is composed of a central core surroundedsuccessively by a membrane, germ cell wall, cortex,coats, and sometimes an outermost layer called exo-sporium.21 In other words, the etch process must pen-etrate through several layers of structural protein be-fore the reactive species can reach and destroy theactual core. Finally, the compositions of small mol-ecules in the spore differ from those of vegetative bac-teria. Spores have much higher calcium and manga-nese contents.21 For example, about 10% of the sporedry weight is composed of dipicolinate anions andcalcium cations. Calcium is concentrated in the core,but significant amounts can also be present in the cor-tex and coats. This high concentration of mineral ions,especially calcium, could lead to the formation of aprotective CaO layer during oxidation by plasma; this,in turn, might retard the etching of deeper-lying con-stituents of the spore. To test this hypothesis, we per-formed preliminary surface analyses of plasma-treated and untreated spores by X-ray photoelectronspectroscopy and by time-of-flight secondary ion massspectroscopy (TOF-SIMS). Because no important dif-ferences—for example, calcium enrichment—were ob-served between the two sample types, this particularinvestigation remains inconclusive. Nevertheless, asignificant organic layer persisted after plasma treat-ment, which resembled the layer before treatment.

A second hypothesis is that the microscopic, ellip-soidal morphology of the spores may help protectthem against plasma etching. To test this hypothesis,we subjected PS microspheres 1 mm in diameter [Fig.8(a)] to the same O2/CF4 plasma etch conditions usedfor spores. As shown in Figure 8(b,c), residues werepresent after t =15 min, whereas a PS film would haveleft no trace.15 However, the observed residues ap-peared to derive from the outer perimeter of the par-ticles [Fig. 8(c)], not from the center. This calls forfurther investigation, now in progress.

Sterilization mechanisms of commercial processes

In contrast to the in-house plasma treatments, ster-ilization by steam, pure EO and Sterrad® did not in-duce an apparent loss of spore material [Fig. 6(d–f)].Indeed, SEM observation revealed that etching did notplay a significant role in the Sterrad-100S® steriliza-

tion system. On the contrary, spores subjected to Ster-rad® appear larger [Fig. 6(d)], which is most likelyindicative of chemical denaturization—for example,oxidation—which can be induced both by the vapor-ized H2O2 (chemical phase), and by other reactive spe-cies during the plasma phase.17,23 However, Krebs etal.11 recently showed that the H2O2 chemical phaselargely dominates over plasma effects in Sterrad-100®sterilization. Regarding spores killed by steam [Fig.6(e)], their outer membranes appear to have been dis-rupted, probably owing to denaturation of their pro-

Figure 8. Scanning electron micrographs of polystyrene la-tex particles: (a) untreated control; (b) and (c) two differentlocations after 15 min exposure to O2/CF4 plasma (p = 80mTorr; P = 200 W; F = 50 sccm; d = 6 cm [CF4] = 15%).

134 LEROUGE ET AL.

teins by hydrolysis.24 EO sterilization, on the otherhand, has little apparent effect on spore morphology[Fig. 6(f)]. This was not unexpected, considering thatthe accepted mechanism is alkylation of specificchemical functional groups of nucleic acids in the cellnucleus.2

CONCLUSIONS

All plasma treatments we investigated proved ca-pable of killing Bs spores, but their relative efficaciesvaried significantly among the various precursorgases. We demonstrated, to the best of our knowledgefor the first time, the particularly high efficacy of O2/CF4 (88/12%) plasma for sterilization, which we haveattributed to its high etch rate. O2/CF4 might thereforebe used to advantage as a gas mixture in new plasmasterilization technologies. Because we have clearlyshown plasma etching to be a key contributor to sporemortality, plasma is also likely to prove effective indestroying prions and endotoxins, which most steril-ization processes fail to inactivate.25 However, sporedestruction by plasma leaves solid residues, whichmay result in adverse biological reactions. More gen-erally, because etching is a nonspecific mechanism,O2/CF4 plasma will also attack the surface of poly-meric biomedical devices being sterilized, resulting insurface modification—for example, ablation, partialfluorination, and oxidation. This may have conse-quences for the biological and physical performance ofthe sterilized devices,26,27 and it must be considered inthe application of plasma to sterilization. This is animportant subject of current investigations in our labo-ratories.

The authors express gratitude to Dr. Y. Deslandes (Na-tional Research Council, Ottawa) for time-of-flight second-ary ion mass spectroscopy studies. The authors also thankDr. G. Czeremuszkin, Dr. M. Latreche, and A. da Silva fortechnical support and many useful discussions; and L. Lefe-bvre, M.-L. Beauchamps, K. Julien, and J. Proulx for provid-ing microbiological data.

References1. Egitto FD, Vukanovic V, Taylor GN. Plasma etching of organic

polymers. In: d’Agostino R, editor. Plasma deposition, treat-ments, and etching of polymers. Boston: Academic Press, 1990.p 321–422.

2. Steelman VM. Ethylene oxide: The importance of aeration.AORN J 1992;55:773–787.

3. Menashi WP. Treatment of surfaces. U.S. Patent No. 3,383,163,1968.

4. Boucher RMG. State of the art in gas plasma sterilization. MedDev Diagn Ind 1985;7:51–56.

5. Lin S.-M. Interaction of bacterial spores with radicals gener-ated by microwave and low-temperature radio-frequency dis-charges. Thesis for doctor of science, University of Texas atArlington, May 1986.

6. Nelson CL, Berger TJ. Inactivation of microorganisms by oxy-gen plasma. Curr Microbiol 1989;18:275–276.

7. Baier RE, Carter JM, Sorensen SE, Meyer AE, McGowan BD,Kasprzak SA. Radiofrequency gas plasma disinfection of den-tal operative instruments including handpieces. J Oral Implan-tol 1992;18:236–242.

8. Chau TT, Kao KC, Blank G, Madrid F. Microwave plasmas forlow-temperature dry sterilization. Biomaterials 1996;17:1273–1277.

9. Lerouge S, Tabrizian M, Wertheimer MR, Czeremuszkin G,Marchand R, Yahia L’H. Etude de l’influence des parametresexperimentaux sur l’efficacite de la sterilisation par plasmafroid. International Symposium on Advanced Biomaterials,Montreal, October 2–5, 1997.

10. Hury S, Vidal DR, Desor F, Pelletier J, Lagarde T. A parametricstudy of the destruction efficiency of Bacillus spores in low-pressure oxygen-based plasmas. Lett Appl Microbiol 1998;26:241–247.

11. Krebs MC, Becasse P, Verjat D, Darbord JC. Gas-plasma ster-ilization: Relative efficacy of the hydrogen peroxide phasecompared with that of the plasma phase. Int J Pharm 1998;160:75–81.

12. Boucher RMG. Seeded gas plasma sterilization method. U.S.Patent No. 4,207,286, 1980.

13. Pelletier J. La sterilisation par le procede plasma. Agressologie1993;33:105–110.

14. Lamontagne B, Wrobel AM, Jalbert G, Wertheimer MR. Large-area microwave etching of polyimide. J Phys Appl Phys 1987;20:844–850.

15. Wrobel AM, Lamontagne B, Wertheimer MR. Large-area mi-crowave and radiofrequency plasma etching of polymers.Plasma Chem Plasma Proc 1988;8:315–329.

16. Lamontagne B, Kuttel OM, Wertheimer MR. Etching of poly-mers in microwave/radio-frequency O2-CF4 plasma. Can JPhys 1991;69:202–206.

17. Jacobs P, Kowatsch R. Sterrad Sterilization System: A newtechnology for instrument sterilization. Endosc Surg AlliedTechnol 1993;1:57–58.

18. Flamm DL, Herb GK. Plasma etching technology: An over-view. In: Manos DM, Flamm DL, editors. Plasma etching: Anintroduction. Boston: Academic Press, 1989. p 91–183.

19. Fozza AC, Kruse A, Hollander A, Ricard A, Wertheimer MR.Vacuum ultraviolet to visible emission of some pure gases andtheir mixtures used for plasma processing. J Vac Sci Technol1998;A16:72–77.

20. Hollander A, Wertheimer MR. Vacuum UV emission from mi-crowave plasmas of hydrogen and its mixture with helium andoxygen. J Vac Sci Technol 1994;A12:879–882.

21. Warth AD. Molecular structure of the bacterial spore. AdvMicrobiol Physiol 1978;17:1–45.

22. Moisan M, Wertheimer MR. Comparison of microwave andradio-frequency plasmas: Fundamentals and applications. SurfCoat Technol 1993;59:1–13.

23. Ikarashi Y, Tsuchiya T, Nakamura A. Cytotoxicity of medicalmaterials sterilized with vapour-phase hydrogen peroxide.Biomaterials 1995;16:177–183.

24. Miller CH. Update on heat sterilization and sterilization moni-toring. Compendium 1993;14:306–331.

25. Steelman VM. Creuzfeldt-Jakob disease: Recommendations forinfection control. Am J Infect Contr 1994;22:312–318.

26. Ratner BD, Chilkoti A, Lopez GP. Plasma deposition and treat-ment for biomaterial applications. In: d’Agostino R, editor.Plasma deposition, treatment and etching of polymers. Boston:Academic Press, 1996. p 463–511.

27. Lerouge S, Tabrizian M, Moreau S, Wertheimer MR, MarchandR, Yahia L’H. Reuse of catheters: Surface modification inducedby gas plasma and pure ethylene oxide sterilization. In: Pro-ceedings of the Society for Biomaterials, San Diego, CA, March1998.

135MECHANISM OF PLASMA STERILIZATION