I debated between Answer or Comment on this, but ultimately I think it is more of an answer - albeit an imperfect answer.
It sounds to me like the primary problem you are trying to solve is preventing material transfer from the balls/cups etc. during the ball milling process. I don't think you necessarily need to change materials, and perhaps could even get away with just adding a PVD or CVD coating to the existing components.
DLC (diamond-like-carbon) coating pops to mind first,
however I think there are many coatings that could potentially serve you well. DLC coatings are very hard and wear resistant (as the "diamond-like" name would suggest). They do not flake or chip off, and while I don't know the end use or the nature of the "no contamination" requirement, these coatings are completely inert to almost all organic and synthetic chemicals, and completely inert to the human body as well.
For this application, I would think a ta-C or possibly a ta-C:H coating could work well. Another DLC that is supposed to be extremely hard and wear resistant, but that I have no personal experience with, is UNCD (ultrananocrystalline diamond).
You may also find more traditional tool coatings, such as TiN or TiAlN could work well for you - I couldn't make a recommendation as to the best coating for your application. I'm not a coatings engineer, just an engineer that has had good experience using these types of coatings for my own applications in the past :)
The main benefit of 440C is it's high hardenability
With a maximum hardness of around 60 rockwell C, 440C is a stainless that can rival many tool steels.
It sounds like you are using this material in the annealed condition, which is questionable to me. This material is typically chosen for it's high hardenability - it's not (to the best of my knowledge) commonly used in the annealed state.
It sounds like you chose this material because it's commonly used in other similar designs; I wonder what condition is this material used in those designs.
If you can get your hands on some 440C components from one of those other designs I would do a rockwell test to see whether or not it's heat treated. I would put money on it being heat treated. Annealed material will probably be in the 20's-30's rockwell C, while heat treated material is going to be in the 50's.
If you do intent do to a rockwell test, try doing in on a component with a flat as spheres can be difficult to test and give inaccurate readings.
For a place to start with heat treat, I find Carpenter's data sheets are usually pretty reliable. Their recommendations for 440C are:
- HARDEN: Heat to 1850/1950ºF (1010/1066ºC); soak; quench in warm oil or cool in air. Hardness will be ≈60HRC. Do not overheat or you will not be able to attain maximum hardness.
- TEMPER: To remove peak stresses and yet retain maximum hardness, temper at least one hour at 300/350ºF (149/177ºC).
If you are looking to heat treat, I would probably start there.
I do not think 316 would be a good choice
As it's a much softer (gummier) material than 440C, and I think it would exasperate the material transfer problem.
Now I said this was an imperfect answer,
because I don't know that it directly answers your question. It offers an avenue to a possible solution, however it is incomplete as you'll need to discuss with a coating expert the exact needs of your application and see what they say.
For example, I have no knowledge of how these coatings perform at cryogenic temperatures, or what coating would work best with the abrasives used in your ball milling process.
I know I have used DLC coatings to solve some unique problems where I work, and they have allowed me to do things I don't think I could have found an alternative for.
I also wanted to say that finding a reputable and reliable coating vendor was the hardest part of adapting this technology for me. As that can be quite a daunting challenge, I offer a suggestion only as a place to start, and with no further endorsement other than I've personally had good experience with Oerlikon Balzers. I do not know the policy on recommending vendors, and I have no affiliation with any vendors - so feel free to edit this answer to remove the vendor name if it violates any policies.
Regardless of who you choose to do business with, I would highly recommend speaking with one of their applications engineers about the specific needs for your application, and seeing what coating they recommend.
Closing thoughts
PVD/CVD coatings are tribological in nature - they affect the interface with other materials, but do not change the properties of the base substrate.
When we talk about bearing design, there is a property commonly referred to as embedability. This basically refers to a bearing materials ability to absorb (or embed within itself) foreign materials.
Without knowing more about the nature of your design, my gut tells me that you would not want this to happen, as it would mean your balls/cups etc. would be embedding some amount of the alloys that you were trying to grind. It seems unlikely to me that this would be desirable. This is another reason I think 316 would be a poor choice.
To reduce embedability, you want a harder substrate.
My recommendation to you would be to heat treat the 440C components first, then apply a PVD coating afterwards, to obtain the best combination of properties for your application. I would definitely talk to a coatings engineer about what coatings will work for your application (e.g. temperatures, material compatibility, etc.)