Getting to the facts about surface roughness and passivation for 316(L) 4

[jari001]

I work for a pharma company and I quickly became involved in discussions about what the L in 316L meant, Ra values, electropolihing, and passivation for the first time. The basic driver for surface roughness is cleanability and the corrosion resistance treatments are to ensure the purified water/water for injection/purified steam do corrode the metal due to their low ion content (i.e. hungry water syndrome).

I still have some lingering questions:

1) Is the only advantage of chemical passivation vs regualr air exposure the decreased time to form the oxide layer?
2) When trying to achieve a certain Ra value, is the mechanical polishing doing the lion's share of polishing while the electropolishing might get you the last 1 micron?
3) If I electropolish, why am I passivating again? This confuses me because electropolishing should passivate at the same time or am I off base?

I believe the pharma industry has accepted certain practices (like what is described in q3) as some form of GxP, but I need to know the facts, Jack!

Thanks

 

[SwinnyGG]

1) Is the only advantage of chemical passivation vs regualr air exposure the decreased time to form the oxide layer?

No.

Passivation is not about forming the oxide barrier- the stainless steel alloy does that on its own. Passivation is about removing surface contaminants, so that the oxide barrier completely coats the entirety of every surface of the part.

2) When trying to achieve a certain Ra value, is the mechanical polishing doing the lion's share of polishing while the electropolishing might get you the last 1 micron?

This is correct. Electropolishing removes a very very small amount of material- if you electropolish a part that is heavily burnished or scratched, you will have a very shiny, heavily burnished or scratched part.

The electropolishing process is the approximate equivalent of the buffing stage of polishing a car- all of the previous polishing steps need to be correctly executed. Any scratches visible before electropolish will still be visible after.

3) If I electropolish, why am I passivating again? This confuses me because electropolishing should passivate at the same time or am I off base?

If you electropolish, you do not need to passivate afterward. Electropolishing provides the same cleaning as the passivation process does.

 

[jari001]

Would scaling or discoloration constitute surface contaminants or do you mean loose material such as flecks of metal or oils?
Also, is rouge considered surface contamination since passivation processes can remove it? (I have to search E-T about rouge separately, because that is another area that seems unclear for no good reason in pharma).

 

[SwinnyGG]

Would scaling or discoloration constitute surface contaminants or do you mean loose material such as flecks of metal or oils?

I mean all of the above. Correctly passivated parts should be free of scale, loose contaminants, and embedded debris from machine tools or other processes.

Also, is rouge considered surface contamination since passivation processes can remove it? (I have to search E-T about rouge separately, because that is another area that seems unclear for no good reason in pharma).

Whether or not you consider rouging to be a problem depends on the specifications of what you're designing- but electropolishing does reduce the rouging tendency in stainless steels, and you've helped yourself out by selecting 316, as it will be less susceptible to rouging than 304.

 

[jari001]

My company has maintenance tasks in place where the clean utilities and process equipment undergo passivation at determined intervals to remove rouge. My thoughts are: rouging is rust (various iron oxides), the iron oxides are more stable than the native SS so we should move to monitoring based passivation rather than just shelling out money because rouging looks bad. I can't even get people to say "rust" because that is somehow worse than rouge and yet they are the same compounds!

My company is a large pharma company, so our design spec guidelines are usually good(those docs say 316L is mandatory for all product contact surfaces that are part of welded assemblies, 304L if not welded), but at my local site there isn't much true understanding about these issues. Unfortunately, my site is the only one in North America so my international colleagues sometimes screen my emails haha!

I also want to clarify one point. My second quote says passivation process removes rouging and you mentioned electropolishing, so am I wrong about passivation not removing rouge?

 

[Compositepro]

I think there is a little confusion about nomenclature here. Successful passivation requires that the surface first be etched to remove contamination and surface iron molecules so that pure chromium is left at the surface. The formation of an impervious chromium oxide layer on the surface is what causes passivation. The chromium will spontaneously create the oxide layer on exposure to air. Since surface etching is the main activity in passivation, it is often referred to as passivation. But technically this is not correct.

 

[EdStainless]

Removing rouge is foolish. the first question is does it matter? Is it getting into your product? At a detectable level? If you have rouge then find out where it is coming from (hint, 90% is from pump impellers) and fix the source. A stable rouge layer may not look pretty but it does not hurt anything.
If the Fe is coming from actively corroding stainless steel then you have another problem.

I know, everyone uses Ra for surface measurements. All that it measures is flatness, not texture. We had a customer who would buy pickle finished 316L tube and then EP it. It would still have an Ra of >70 microinches but when you looked at samples under an SEM at 100-500x you could not see any surface features. The micro crevices are what you are trying to eliminate.

The selective dissolution of Fe from the surface of SS during passivation is minor. A passivation is cleaning, and it should never etch the surface. If there is attack then you are doing things wrong. After all if it does not come off of the surface in nitric acid will it come off in your process?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[jari001]

My site doesn't have a problem with rouge per se, I think we don't fully understand the phenomenon and we are just accepting overly cautious maintenance plans. I agree that rouge intrinsically doesn't jepordize product safety and quality so I want to influence my teammates to consider a monitoring (conductivity monitoring?) based approach to our maintenance activities. However, QA is always the final word and if they aren't confident they can explain the science to an auditor (a function of how well the engineers can communicate technical information), they'll tell me to move heaven and earth and remove the rouge.

Concerning surface roughness, I used the below source to understand what the profilometer measurements mean. Assuming the measurements taken are the arithmetical mean, I would think that stark surface features of the metal is directly impacting the flatness.

https://my.misumi-ec.com/pdf/tech/press/pr1167_1168.pdf

Thanks for the engagement everyone!

 

[EdStainless]

Profilometers measure the 'wave and lay' of the surface, that is the title of the ANSI spec.
If you read the standard you will see that below 25uin (0.6um) the values are largely just noise.
I have customers that insist that 10uin is 'smoother' than 15uin, but when you examine and measure them optically they may be identical or they may be completely different, but you can't tell from the profilometer values.


= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[tbuelna]

Electropolishing improves surface texture/roughness characteristics by removing material from the tiny surface "peaks" and rounding their sharp edges. But the tiny surface "valleys" between the peaks are largely unaffected. With materials like 316 stainless that are "gummy" and tend to smear when machined, one nice thing about electropolishing is that the process is effective at removing any smeared material left on the surface from machining operations. This would seem to be an important consideration for components that require sterilizing.

 

[EdStainless]

Of the dozens of companies that I worked with over the years there only a few that really cared about the micro-texture of the surface and realized that cleanability was more related to the texture rather than Ra.
We had many customers that required us to mechanically polish tubing, even when the as-manufactured surface met the Ra. And while the mechanical polish lowered the apparent Ra, it created millions of micro crevices that did not help.
Small peaks and tears are to be expected in any mechanically polished surface. If there are laps or smears then the process is very poor should not be used. This is why buffing a surface to make it 'smooth and shiny' is so bad. You smear a lot of surface metal when buffing and the crevices formed are deep and tight making them impossible to clean and a great place for corrosion to initiate.
The ability of EP to actually improve a surface depends on the process parameters and controls (V, A, flow, temp, ...). Unless someone had a documented process, with good controls, and the before and after SEM pictures to back it up then they may be doing nothing but making it shiny, and not really removing the microscopic surface texture at all.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[jari001]

I don't see how surface roughness doesn't correlate with the micro-texture of the surface. In my mind, the more features the surface has the rougher it would be. From NIST (link below), I took profilometers to be measuring the depths of the individual valleys and the heights of individual peaks. Figure 2-1 in the NIST guide provides a pictoral definition of the terms wave and lay, which to me are the coarse features of the surface.

Are those features (individual peaks and valleys) well below the micron scale? Usually 0.22 micron filters are used for sterile filtration, so if my surface can't harbor particles and bugs in the tenths of microns, then the wave and lay measurements should be enough to gauge cleanability of the surface. Otherwise, how am I getting away with NOT getting SEM images Oh boy...

http://www.nist.gov/calibrations/upload/89-4088.pdf

 

[SwinnyGG]

From your posts I think you and Ed aren't far apart in your views- you are just drilling down into the limitations of using Ra as a controlling measurement.

 

[EdStainless]

Jari, you said it yourself, the bacteria that you care about harboring are going to live in crevices that are an order of magnitude smaller than what your profilometer will measure. The profilometer can't measure crevices unless they are larger than the tip.

You have to separate systems that all EPed, and those that are just mech pol. In EP systems the cleanability will be great weather the surface Ra is 0.2u or 1.2u. There won't be any crevices tight enough to trap contaminants and prevent cleaning. Smooth open features that are 100 or 1,000 times the size of contaminants are not a problem, even if your profilometer tells you that it is 'rough'.
In a mech pol system it is a bit different. Once contaminated these can be a real problem to clean.
But in either case if the system is kept clean then it will be fine, and most plants are actually very good at this.

How do you quantify cleanability? Ever done tracer tests on your systems? I have seen a few done, with varying results.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[jari001]

We have specs for product contact surfaces to have a Ra value <= 0.5 micrometers and I am able to produce measurements in this realm with the profilometer we have. Now, the reference to ANSI and the idea that measurements lower than 0.6 micrometers will be more noise than anything else is new to me.

We electropolish all metal surfaces and equipment since 99% of it is 316L SS and we can do so. However, I don't understand why EPed surfaces retain good cleanability. I imagine the spiky areas have a higher current flux and thus more energy goes to these areas and facilitates the mass transport limited dissolution. However, the preferential leveling of these areas doesn't creat wider creavices from existing ones, it just brings down the roughness by reducing the positive displacement from the level. Am I missing something?

As I understand it, my company's stance on clean validation is to perform cleaning performance qualification with swab tests and other tests if needed and then validate that cleaning method if all results pass.

 

[mrfailure]

I was wondering what you mean exactly when you use the term "rouge"? I tend to think of it as iron oxides that get generated under fretting that discolor due to heat generated by the two surfaces rubbing together. Do you mean a loose-oxide phenomenon, or do you really mean coloration of the thin adherent oxide film that is responsible for passivation on the surface?

 

[EdStainless]

In high purity systems it is not uncommon to find polished surfaces that turn red-orange.
It is almost never from the corrosion of that stainless surface, it is nearly always transported Fe from other places in the system. When the Fe in solution gets someplace where temperature, pressure, pH, velocity, gas content, or some other characteristic changes it will form insoluble iron oxide ans settle on the surface.
When you get these ultrafine particles on a very polished surface they cannot be wiped off (sometimes you can get a little to wipe). The only way to remove them is with acid.
In most systems they only bother the Quality people and not the product.
Over 90% of the cases that I have seen the source of the metal was a pump impeller (case 316L is not good enough), with valves and spray ball being nearly all of the others.

In high purity steam systems you will get dark oxides forming. These are very stable and should never be touched.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube