Films can be created by a variety of methods. In this chapter I will describe the ones that are familiar to me, in order of complexity. For all methods of vapor deposition, wavelength stability over time is important; it may be achieved with the careful selection of a starting material and, possibly, some processing before adding the chemicals to the coater.
2.1 Vapor Deposition Methods
2.1.1 Thermal deposit
For many years I have been making filters that have remarkable wavelength stability. The use of zinc sulphide or zinc selenide for high-index films and mixed fluorides or thorium fluoride for low-index films will produce stable bandpass filters when the substrate has been heated to approximately 110˚C. The chemicals are heated to deposit in stainless steel crucibles after pressing the materials into the crucible to improve density. A tungsten spiral filament is positioned over the crucible, and a few amperes at 115 V transformed to 5-20 V will heat the surface to evaporation. The sulphides sublime, and the fluorides melt. Other methods of heating the chemicals can work just as well. After a filter has been deposited, the stability can be guaranteed by heating the filter to approximately 250˚C for a few hours. If the conditions for a stable film have been met, the wavelength should not have shifted more than 1-2 nm. This baking provides a densification of the films. For filters manufactured with cryolite and other mixed fluorides, the best condition for use and storage is in an epoxy medium. Films will survive in air, but they are easy to scratch. Also, if the substrate was not perfectly clean before deposition, there will be a possible pinhole problem over time. If a surface has mobility, there is an opportunity for trouble.
The thermal coefficient of expansion for thermally deposited films cannot be reduced to zero, and this limits their use even though the films are simple to deposit compared to other methods of deposition. The advantage of these films is lack of stress for certain combinations of materials. I have found that the reflective metals are best deposited by the resistance sources; the films are the coldest and the smoothest in a normal deposition sequence.