Essential Basics of Reactive Sputtering

Sputtering is a standard technique for Physical Vapor Deposition (PVD), one of the methods of producing Thin Film Coatings. Basic Sputtering uses a target of whatever pure material is desired, and an inert gas, usually argon. If the material is a single pure chemical element, the atoms simply come off the target in that form and deposit in that form. But it is also possible to use a non inert gas such as oxygen or nitrogen either in place of, or (more commonly) in addition to the inert gas (argon).

When this is done, the ionized non inert gas can react chemically with the target material's vapor cloud and produce a molecular compound which then becomes the deposited film.
For example, a silicon target reactively sputtered with oxygen gas can produce a silicon oxide film, or with nitrogen gas can produce a silicon nitride film.

However, it is also possible to use a non inert gas such as oxygen or nitrogen either in place of, or (more commonly) in addition to the inert gas (argon). When this is done, the ionized non inert gas can react chemically with the target material vapor cloud and produce a molecular compound which then becomes the deposited film.

For example, a silicon target reactively sputtered with oxygen gas can produce a silicon oxide film, or with nitrogen gas can produce a silicon nitride film.
There are numerous variables to be addressed in this process.

One important consideration is the distribution of the reactive gas inside the chamber. Whereas the inert gas atoms can be re-used - ionized, neutralized, and ionized again in the plasma - the reactive gas is chemically consumed to form the desired product compound. In order to produce a consistently reacted film, it is often necessary to include some specific physical means of injecting this gas reasonably uniformly through the entire plasma around the cathode. The typical means of doing this is a "gas ring" surrounding the cathode.

This would be a tube or channel with small holes spaced along it, each allowing a portion of the gas to bleed into the plasma at intervals along the perimeter of the target. Since the reactive gas does ionize and get accelerated into the target by the electrical field, it does sputter material off the target as does the inert gas.

In fact, it is possible to use only the reactive gas, and some users have processes which run that way - 100% reactive gas, no argon. But it is more common to use a combination of inert and reactive gas, and the partial pressure of reactive gas is one of the variables that will influence the stoichiometry and characteristics of the molecular film that is produced. It is also possible for some of those reactive gas ions to react chemically with the atoms on the surface of the target without sputtering them off of the target. This is referred to as "poisoning" the target surface - changing it from a plate of pure original material into one which includes some percentage of already reacted molecules.

And those molecules will behave differently in terms of conductivity, sputter yield, and so forth.

For this reason, it can be necessary to periodically recondition the surface by sputtering off this top layer with just the inert gas to get back to original material.

If the original target material is suitably conductive, DC sputtering is possible, and the use of a pulsed DC power supply can help reduce the need for this regeneration step, which is otherwise an interruption of normal processing. Molecular films can also be produced by standard (inert) sputtering of molecular material targets.

For example, the previously mentioned silicon oxide or silicon nitride can be argon sputtered from targets made of those molecular materials.
But ions slamming into a target with sufficient energy to sputter whole molecules off the surface are also capable of tearing some of those molecules apart.
For this reason, even if reactive sputtering is not the basic process, it can still be useful to include some amount of background reactive gas to restore stoichiometry for these dissociated fragments. One of the more useful accessories to have on a sputter system meant for reactive sputtering is a Residual Gas Analyzer (RGA) or Process Gas Analyzer (PGA) in order to monitor the actual partial pressures of gas species involved in the process.

These can then be controlled by means of gas Mass Flow Controllers (MFC's) along with a pumping throttle valve and suitable control instrumentation. Reactive sputtering can be a dynamic, rather than static, environment, with these partial pressures being a key parameter for process optimization.
As with all such things, there will be details to work out in order to establish a good working process for your end product, but Reactive Sputtering can be a very useful tool to have available for creating molecular films, especially if there is a need to tinker some with atom ratios (stoichiometry) in the film in order to control properties such as resistivity or index of refraction.