What kind of ion sources is the best for ion beam milling?
You have to evaluate the following parameters:
The ion source should provide the flow of compensated plasma with ion current density about a few milliamperes per square centimeter and with energy of ions less than 200 eV. In this case you'll get technologically good sputtering rates with low surface defects concentration. The dimensions of beam cross section depends on the dimensions of an optical part to be milled (see Section "Ion Beam Milling"). It is better to use griddles ion sources to minimize possibilities of an optical part surface contamination with particles sputtered from the ion source parts. If you are going to realize processes with the moving ion source, think about its design because all wires and pipes should move too. Therefore it would be better to use low voltage ion source with minimum electrical wires and no water cooling . In my opinion the Hall accelerators with beam aperture about 30 - 50 mm should be the best.
Does ion beam milling contaminate the optical surface?
Sure it does. Some parts of any ion source (grids, electrodes, walls of discharge chamber, etc.) can be sputtered by created ions, but in well designed one this possibility is minimized. In principle you can separate a ion flow by ion masses, but it is not technological solution (too complicated!). The question is: What contamination level is accepted for your goals? Usually you can accept contamination if they don't change the optical parameters of the surface. In my experience I didn't run into this problem: optical parameters of milled surfaces were OK. But I saw that the walls of accelerating channel of my Hall accelerator became thinner and thinner with time (long time - years!) due to the sputtering. Of course we sputter surface with the sputtering rate that is much more than the rate of contamination by particles sputtered from the ion source parts, but these particles can deposit in the areas of surface outside of beam sputtered area. On other hand my source accelerating channel had been made from ceramic, so the deposited layer could consist of a mix of transparent oxides and nitrides with low refraction index. By the way I know about one investigation when after the optical glass surface sputtering by Hall accelerator the contamination of surface with practically all materials of ion source has been obtained. However it has been obtained only with the help of Auger and SIMS spectrometry methods practically in the quantity less than one mono layer. This contamination practically didn't change the surface optical parameters.
Another kind of contamination may be much more dangerous if you are working with diffusion pump and these is an oil vapor in chamber. In the presence of a plasma the oil vapor may be polymerized on the surfaces inside the chamber including the optical part surface. You can seethis contamination as a very thin but very hard light yellow layer. It may be removed only by mechanical polishing. Probably, it is something like diamond-like coating. So better use vacuum system made on the base of turbomolecular pumps or take strong measures to provide oil free vacuum in the vorking chamber
.What is the best position and motion of an ion source and part should be milled?
I think the best position is when the optical part is rotating or turning at given angle around it axis of symmetry and the ion source is moving from the axis of part rotating to it periphery. Of course here is a problem how to move the ion source with all its wires and pipes in vacuum , but it is not a big problem. For big optical parts it is easier to realize the linear motion of an ion source with a weight about a few pounds than move an optical part with a weight more than hundred pounds. Second reason is that if you are going to move an optical part the overall dimensions.
Because of the big optical part weight it looks better if a part will be put onto a part holder and an ion source will move above a part surface (something like shown on my main page). In this case you can unload a part on a holder.
Should we expect the re-deposition of sputtered material back on the surface?
It depends on vacuum conditions in your chamber. If the pressure in vacuum chamber is about (2 - 4)x10-2 Pa, the free path length of sputtered particles is about 50 cm - 100 cm. Therefore the possibility that these particles will change the direction of their propagation back to the sputtered surface due to collisions with other particles is very small.
Can we use the ion beam milling for aspherization?
In general you can. But it depends on required deviation of your aspherical surface from a nearest sphere. Usually this value is about ten micrometers or more and I am not sure you will be able to remove that quantity of material and save the surface quality. Besides it may be very long time process for the parts with big dimensions. I think the best way is to use conventional technologies for rough pre-aspherization and then the ion beam milling as super finish retouching.
How can I increase the sputtering rate?
You can increase the beam power but you must be very careful because of possible part overheating. The usage of chemically active gases (freons, for example) allows to increase sputtering rates in a few times, but at some countries freons are forbidden for usage due to the problems with the ozone layer around the Earth. If you will look for other chemically active gases don't forget to take into account that some of them can destroy parts of your equipment due to a strong corrosion of metal parts and a destruction of vacuum oils. I obtained this effect working with freons.
You were going to add the Section "Ion Beam Assisting" and haven't written it. Why?
Yes, I was.... I refuse of this idea because of the following reason: All the written here stuff is a kind of a general principles which my be used as the fundamentals of any work. In case of ion beam assisted coating deposition (IAD) the process parameters very strongly depend on the given task and equipment to be used. And I can't guarantee that you'll obtain the same or just good results using my experience for your goals. Sometimes I couldn't repeat my former results on another equipment (at least it took a lot of time to repeat them....).
If you have any questions about IAD please don't hesitate to ask me. I'll be glad to help you to solve your private problem, if I will be able....