We make OLED deposition systems, glovebox and otherwise. Most are very simple and either deposit the organics or the interconnect. Because of the fragility of the devices, patterning is sometimes done by deposition through stencil masks – that is, lithographic masks with holes, placed in proximity to the OLED substrate. This ancient, low-resolution technique is sufficient for test devices, although production might be a different matter.
Anyway, to make a multi-layer device, users want an in-situ mask changing mechanism, and they also need some means of cleaning the masks. As with everything else in life and vacuum, there are easy ways and hard ways. To make it possible to change masks under vacuum with the absolute minimum of gear, we came up with a very simple carousel device operating from a single feedthrough. Of course, every user seems to have a different mask size and substrate size, so we’re not exactly going to mass-produce this item. We are, however, beginning to run into mask cleaning and handling questions, so here are a few ideas.Stencil masks are limited in resolution by three main factors: the mask thickness, the evaporation source size, and the closeness of contact to the substrate. For photo etched metal masks (we use stainless) the mask thickness is a couple of mils, and anything over about one cm lateral extent may require a stiffener layer for good contact. Mechanical mask registration and the other factors combine so that features smaller than a few mils, registered on similar features, are difficult. This is no concern for large area test devices, but some of you are making displays this way, displays that can’t have point defects.
Depositing some thick organic on a metal mask means that the mask is slowly obscured, becomes ’sticky’ and may change dimension. Cleaning the mask out of the system is often just a matter of a wet dip and an ultrasonic wash. Cleaning in place is best accomplished with a plasma, and this is a good idea as a final clean in any event. Depending on the material, the plasma may be glove box residual atmosphere, such as argon, or an introduced chemistry. The plasma can be created as a low or high frequency glow discharge using electrodes external to the substrate, or the substrate holder itself. The cleaning plasma can also augment the chamber or chamber liner cleaning cycle. Right now we’ve got a batch of nickel masks that are for relatively large displays, and they’re extremely difficult to keep clean in the conditions prevailing in the user’s lab. In situ cleaning is the only way to get any yield at all.
Give us a call if patterning is proving to be a questionable process, or just if you have something special that you need designed in this area.
