There are a lot of reasons why there may be an arcing problem during the plasma deposition process during sputtering. During the sputtering process a gas, typically argon unless a reactive deposition is required, is transformed from a neutral state or insulating state to a charged or conductive state. This is caused by stripping off an outer electron, thus leaving a positive charge on the molecule. This positively charged particle (ion) can then be directed toward the sputtering target via an electrical and/or magnetic charge which has an overall negative charge. In most cases it is possible to control this conductive plasma with the ions striking the atoms of the target, breaking the atomic bonds and propelling the dislodged atoms toward the substrate to produce high quality resultant films but occasionally the process breaks down and arcing is formed in the plasma. Arcing is a high power density short circuit from the plasma to ground potential. This is the result of the working gas actually becoming more active, i.e. more electrically conductive, than is normally required for thin film deposition. There are many causes for this overly active state to occur, some are mechanical, some are electrical and some are merely topographical manifestations or imperfections on the target surface itself; such as pits, voids, cracks, or protrusions. All of these factors increase the power density within the plasma, locally, and result in an arc. One of the most common, and one of the easiest to rectify, is when the arc occurs between the target surface and the dark space shield.
Since this is one of the easiest to explain and one of the easiest to fix, of course I have chosen this one to discuss. The dark space shield is typically a simple metal ring or some form of an adjustable protrusion that extends beyond the target surface and at some pre determined distance equally spaced around the outer perimeter of the target. The dark space shield should be at ground potential. The target, during the deposition process, has a high electrical potential on the surface. If there is any stray material bridging the gap between the target and the dark space shield, such as accumulated material that has fallen off the shutter, target, shield, chamber wall, etc. then it will cause a short circuit and/or arcing within the plasma. This is very detrimental to both the target and the deposition system in general. If this occurs it will be necessary to open the chamber and clean the debris away with either sand paper or a 3M Scotch Brite pad.
The dark space shield in the deposition system should be adjustable. Adjustable laterally as well as vertically. When installing a new target, make certain that the gap between the outer perimeter of the target and the dark space shield is uniform in distance around the whole circumference of the target. The gap distance is typically designed to be approximately 2-4mm or so. Any variation in the gap distance from one position to another will change the impedance within the network and possibly arc over to ground potential wherever the shortest gap distance exists. Additionally the edge of the dark space shield that faces toward the sputtering target should either extend slightly beyond the target surface or at least be even with it. This creates the actual “Dark Space” or more specifically an area in the vicinity of the target edge where no plasma exists during the deposition process. This prevents deposition from the target, as well as re-deposition within the plasma, from building up along the edge of the dark space shield or the edge of the target. Both of these detrimental scenarios will eventually cause the gap distance to diminish and eventually short out the target or arc over to ground potential. Care should be taken to check this height after installing a new target as the target thicknesses may vary from one to another. This may especially be true if some targets are bonded to backing plates and some are not.