On glow-discharge sputtering of iron and steels in a commercial deposition plant
P. Kringhøja , C.V. Budtz-Jørgensena , J. Find Nielsen a , J. Bøttiger a,U , S.S. Eskildsen b, C. Mathiasen b
b a Institute of Physics and Astronomy, Aarhus Uni¨ ersity, DK-8000 Aarhus C, Denmark Tribology Centre, Danish Technological Institute, Teknologiparken,DK-8000 Aarhus C, Denmark
Received 29 August 2000; received in revised form 12 October 2000; accepted 12 October 2000
Abstract The sputtering rates of iron ﬁlms and various steels exposed to argon hydrogen pulsed-DC glow discharges in a commercial plasma-assisted chemical vapour deposition plant have been studied experimentally. The sputtering rate of iron was obtained from measurements ofthe ﬁlm thickness before and after glow-discharge exposure, using Rutherford backscattering spectrometry, and the sputtering rates of the steels were obtained from shifts of positions of implanted Xe markers in Rutherford backscattering spectra. Through anode-catcher experiments, the dependence of the sputtering rate on the oxygen partial pressure was measured. To investigate any inﬂuence of thebackground oxygen pressure on the plasma Žthe cathode ion bombardment., energy spectra of the ions hitting the cathode were also recorded. With ﬁnite oxygen background pressures, iron sputtering rates much smaller than the ones corresponding to sputtering yields of iron oxides were measured. In fact, due to these low rates, it was only possible to establish upper limits to the iron and steelsputtering rates in glow discharges in the commercial chamber. 2001 Elsevier Science B.V. All rights reserved.
Keywords: Argon; DC pulsed; Energy distribution; Glow discharge; Iron sputtering; Oxygen partial pressure
1. Introduction Sputtering, i.e. the erosion of a solid due to bombardment with energetic ions, has been investigated for more than a century w1x. Especially in the last four decades,extensive experimental w2x and theoretical studies w3,4x have been carried out. However, at the atomic scale many details are still uncertain. The sputtering yield, deﬁned as the mean number of atoms removed from the surface per incident ion, is the most widely measured quantity. Reactive Žor chemical. sputtering, e.g. is poorly un-
Corresponding author. Tel: q45-8942-2899; fax: q45-8612-0740.E-mail address: firstname.lastname@example.org ŽJ. Bøttiger..
derstood w2x. During reactive sputtering, various elements, which are being implanted into the surface andror chemically react with it, change the surfacebinding energies. These energies enter inversely in the sputtering yield. The inﬂuence of surface topography and absorbed layers is also not well understood. In a glow discharge ŽGD., dueto the formation of the cathode sheath with a large electrical ﬁeld, energetic ions bombard the cathode and cause sputtering. This is being used intensively in commercial plasmadeposition systems for cleaning substrates in-situ before deposition of ﬁlms. In a recent experimental study, Budtz-Jørgensen et al. w5x measured the physical sputtering of gold surfaces by argon hydrogen pulsed-DC glowdischarges. The energy distributions of ions hitting the cathode were
0257-8972r01r$ - see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S 0 2 5 7 - 8 9 7 2 Ž 0 0 . 0 1 1 0 7 - 5
P. Kringhøj et al. r Surface and Coatings Technology 137 (2001) 277 283
also measured, and it was shown that from these the sputtering rates could be calculated, using knownsputtering-yield curves. It was also found that the largest contribution to the ion sputtering originated from ArHq ions. Using the above method to evaluate the physical sputtering, Budtz-Jørgensen et al. w6x have also studied the contribution of chemical sputtering to the erosion of Al 2 O 3 in a argon hydrogen pulsed-DC GD. Heller w7x has measured the sputtering rates of iron and cobalt targets in an...