Surface and Coatings Technology 174–175(2003) 570–573

0257-8972/03/$ - see front matter� 2003 Elsevier Science B.V. All rights reserved.doi:10.1016/S0257-8972Ž03.00630-3

Plasma diagnostics in a double inductively coupled source(DICP) forplasma sterilisation

P. Messerer*, B. Boenigk, G. Keil, P. Scheubert, P. Awakowicz

Institute for Physics of Electrotechnology, Munich University of Technology, Arcisstr. 21, Munich D-80333, Germany

Abstract

As a response to the need for sterilising thermolabile implant materials like polylactide or ultra-high-molecular-weight-polyethylene(UHMWPE) without toxic or water containing liquids or gases, a new kind of plasma reactor was developed. Inorder to achieve homogenous sterilisation efficiency, this reactor is equipped with two flat rf coils, one on top and one on bottom.Based on harmless and well-known gases like argon, hydrogen and nitrogen, good sterilisation results withBacilus subtilis testspores have been obtained. To investigate the different plasma mechanisms necessary for the reduction of germs and spores, theplasma was analyzed by Langmuir probe(APS3) measurement, mass spectrometry and simulation. The design of the electro-polished stainless-steel chamber and the coil configuration was also based on simulation. The in-house plasma model consists ona hydrodynamic code for transport behavior and an electrodynamic part for rf couplingwScheubert et al. J. Appl. Phys. 90(2)(2001) 587–598x.� 2003 Elsevier Science B.V. All rights reserved.

Keywords: DICP; Model; Sterilisation

1. Introduction

Numerous advantages of inductively coupled plasmas(ICP) are well known w2x. The high electron densityand the low impact power of ions make them ideal formany applications in surface technologies. Intensive UVlight in combination with high radical flux densitiesreveals fast process rates, i.e. for etching or thin filmdeposition. On the other hand, the low kinetic energy ofthe heavy particles is ideal for sensitive materials. Upto now, these plasmas needed to be homogeneous intwo dimensions, for example, to process wafers insemiconductor production. Nowadays, new applicationsare of special interest. It is relevant for thermolabilematerials in food packaging or biodegradable materialsfor medical implants to treat three-dimensional objects.That is the reason why the double inductively coupledplasma reactor(DICP) was developed with reasonablehomogeneity in axial and radial direction.In order to find correlations between the parameters

and sterilisation efficiency of various plasmas for differ-ent test spores and germs(Bacilus subtilis, B. stearoth-

*Corresponding author. Tel.:q49-8928923126; fax: q49-8928923134.

E-mail address: messerer@tep.ei.tum.de(P. Messerer).

ermophilus, A. niger and others), the various plasmaswere investigated by a Langmuir probe system(APS3), a plasma monitor(EQP 500, Hiden Analytical) andan electrodynamic-hydrodynamic model.

2. Setup of the DICP-reactor

The cylinder symmetric stainless-steel chamber(diameter 400 mm, height 200 mm) is electro-polishedand equipped with two quartz plates, one on top, theother on bottom. Two rf coils mounted above the topand below the bottom serve as rf antennas for rf powerincoupling. Both coils consist of three concentric silver-coated copper rings, which are connected to an automi-sed matching network(Aurion Anlagenbau GmbH).This so-called match box combines two conventionalcapacitive power splitters to drive each rf antennaseperately. Chamber design, coil configuration and thelayout of match box components have been carried outby simulation. The whole setup is controled by a Lab-View software to improve reproducability and reliability.

3. Simulation

To describe the transport processes in the DICP,balance equations derived from Boltzmann’s equation

571P. Messerer et al. / Surface and Coatings Technology 174 –175 (2003) 570–573

Fig. 2. Comparison of simulation and measurement.

Fig. 1. Simulated amount of electromagnetic field strength of an ICP and a DICP with argon at 10 Pa and 600 W input power.

were used. The tranport of mass, momentum, and energyis described by assuming Maxwellian energy distributionfor all particle species. The discharge including the coilswas considered as cylinder symmetric, while the neutralgas density was assumed to be spatially constant. Onespecies of positive ions is included, whereby the balanc-es of mass and momentum were used. For electrons,conservation of mass and energy are considered. Electricinteraction between electrons and ions is given byPoisson’s equation. The complete model as well as theelectrodynamic part is described in Ref.w1x. Theimprovement of the DICP is given in Fig. 1, where theamount of the induced electric field in an argon plasmafor both, DICP and normal ICP sources is compared.In Fig. 2, spatial profiles of electron density in an

argon plasma measured by probe are compared tosimulation. As the gas temperature is an input parameterfor the model, it was determined by optical emission ofthe rotational band of the N second positive system,2

approximately at 380 nm. For this purpose, a smallamount of nitrogen was added to the plasma(-0.5%).The dependency of different gas temperatures is alsoshown in this figure.

4. Langmuir probe measurements

The in-house built Langmuir probe system APS3 isdescribed in detail in Ref.w1x. By measuring the probecurrent at plasma potential, the electron density wasevaluated. The mean electron energy was calculatedfrom the electron retarding current in the range offloating and plasma potential. Since the efficiency ofthe whole rf circuit from the generator to plasma isunknown, the electron density of the chamber’s middle

axis serves as the absolute value of rf power input forthe model. As a result, the efficiency covers a broadrange from some percent in the low-pressure regionbelow 1 Pa up to 50% in the region around 50 Pa.Fig. 3 shows measured electron density profiles from

the wall to the middle of the reactor at different rfpower values. Operated in argon at 10 Pa, the densityincreases with power and forms a Bessel-like electrondensity profile.As a result, by increasing the pressure above 50 Pa,

the electron density profiles start to flatten in the middleand a torus like shape can be observed(see Fig. 2).In order to enhance the sterilisation efficiency, meas-

urements with different gas mixtures of argon andnitrogen, or argon and hydrogen have been performed.

572 P. Messerer et al. / Surface and Coatings Technology 174 –175 (2003) 570–573

Fig. 3. Measured electron density profiles for argon at 10 Pa andvarious rf power input.

Fig. 5. Temporal reduction ofB. subtilis.

Fig. 4. Atomic nitrogen(left) and the sterilisation efficiency(right) show a maximum near a gas composition of 100:5 Ar:N .2

For the optimum gas mixture of argon:hydrogens200:10, the electron density drops to values below 1011

cm . As long as the gas mixture is kept constant, they3

sterilisation efficiency is proportional to the measuredelectron density. By varying the mixture, it does nolonger correlate to the electron density.

5. Mass spectrometry

The plasma monitor EQP 500 is a quadrupole mass(up to 500 amu) and energy analyzer equipped with achanneltron detector. Ion fluxes, ion energies and num-ber densities of radicals have been investigated. In orderto detect radicals from the plasma, the filament temper-ature of the ion source in the instrument has to beadjusted. Number densities of nitrogen atoms have beenmeasured by ionisation threshold mass spectrometrywhich is described elsewherew3x.

6. Sterilisation results

The observed intensity of atomic nitrogen in Fig. 4shows a correlation between the reduction of test spores(B. subtilis), and the amount of atomic nitrogen. Thesame holds for hydrogen atoms in argon hydrogenmixtures. For sterilisation, we found an optimal gascomposition of 5 parts of hydrogen to 100 parts ofargon. The best results for a mixture of nitrogen inargon were at the same proportion. The advantages ofhydrogen over nitrogen may be contributed to thecontinuum of UV light w4x of the hydrogen spectrum,which has no counterpart in nitrogen.In Fig. 5, the temporal reduction ofB. subtilis test

spores in an argon hydrogen plasma is shown. TheUHMWPE and polylactide substrates have been contam-inated by spraying, approximately 10 spores onto one6

item. Each test has been repeated three times to improvethe statistical variation of the result.

573P. Messerer et al. / Surface and Coatings Technology 174 –175 (2003) 570–573

A reduction of spores of six orders of magnitude inseveral minutes can be achieved at a rf power input thatdoes not harm sensitive materials, like polylactidescrews or hip joints made of UHMWPE. A modificationof the surface of UHMWPE can be achieved to improvethe wear resistance by an admixture of hydrogen and anadjustment of several external parameters. Even moresensitive materials can be treated by using a pulsedplasma. First tests with thin plastic films have beenmade and are still under investigation.

7. Conclusion

In this paper, a double inductive coupled plasmasource(DICP) is operated at argon, nitrogen and hydro-gen for the purpose of plasma sterilisation. First, diag-nostic results of the DICP correlated with sterilisationtests give an idea of the most important plasma ingre-dients for plasma sterilisation. In order to understandthe physical behavior of the DICP, Langmuir probemeasurements for spatial electron density profiles havebeen carried out. With mass spectrometry, relative con-centrations of atomic nitrogen and hydrogen were meas-ured. First material tests revealed a low damaging of

the treated polylactide and UHMWPE substrates. In thenear future, ion and neutral beam experiments will beset up to obtain sterilisation results with isolated effectsof the respecting plasmas.

Acknowledgments

The authors would like to thank the Dr JohannesHeidenhain Stiftung for funding parts of this project.The investigations were mainly supported by the federalministry of education and science(BMBF) under thenumber of 13N7681.

References

w1x P. Scheubert, U. Fantz, P. Awakowicz, H. Paulin, Experimentaland theoretical characterization of an inductively coupledplasma source, J. Appl. Phys. 90(2) (2001) 587–598.

w2x Electron kinetics and applications of glow discharges, in: U.Kortshagen, L.D. Tsendin(Eds.), Chapter Optical Character-isation of RF Inductively Coupled Plasmas, Plenum Press,New York, 1998, p. 489.

w3x P. Pecher, Thesis, Univ. Bayreuth(1997).w4x J. Feichtinger, Proceedings of the Eight International Confer-

ence on Plasma Surface Engineering(2002), to be printed.