I. Safi, A novel reactive magnetron sputtering technique for producing insulating oxides of metal alloys and other compound thin films, SURF COAT, 135(1), 2000, pp. 48-59
Problems associated with reactive magnetron sputtering from elemental (i.e.
non-compound) targets have been successfully solved in this work. The elem
ents of this achievement are: (i) the use of mid-frequency (i.e. 40 kHz) AC
power in the floating mode, between two magnetrons, allowed the reactive s
puttering process to be are-free and hence eliminating the undesired effect
s of arcing in reactive sputtering such as driving the process to become un
stable, creating defects in the films and reducing the target lifetime. (ii
) The combination of DC and mid-frequency AC power in a novel way using a f
ilter to protect the DC power supply from the AC one (or the independently
DC powered magnetrons method) permitted the composition of the produced fil
ms to be easily and independently manipulated by varying the magnitude of p
ower applied to each magnetron. (iii) The use of very fast feedback methods
to automatically control the admission rate of oxygen into the sputtering
chamber (i.e. plasma emission monitoring or voltage control) allowed the st
oichiometry of the deposited films to be independently controlled. This als
o allowed the deposition rate of the sputtered films to be high. (iv) Sputt
ering from two magnetrons made the production of alloys or multi-element co
mpounds, which are either difficult or impossible to be formed from single
targets, an easy task. (v) Substrate rotation enhanced atomic level mixing
of the him constituents. The stoichiometry of the film was controlled by pl
asma emission monitoring or voltage control on one magnetron, and dopants w
ere added by sputtering from the other magnetron. This means that the forme
r magnetron served two purposes; the first was to sputter metal and oxidise
it, and the second purpose was to oxidise the metal sputtered from the oth
er magnetron. This novel technique opens the door wide for investigating vi
rtually all potentially promising thin oxide films. Using this technique, a
large range of alloy-oxide films was deposited at high rates. In fact, the
independent control of both the metallic composition and stoichiometry was
very valuable in identifying the optimum properties of these films. That i
s, giving transparent films of different refractive indices for optical app
lications. Furthermore, such a technique may also be capable of investigati
ng other types of thin films (e.g. hard coatings, semiconducting films, sup
erconducting films, etc.). (C) 2000 Elsevier Science B.V. All rights reserv
ed.