Mobile Phase Filtration

[Eric Moore]

Hi, all -

A colleague (non-chromatographer) asked me the following questions today, "Why do you vacuum filter your mobile phase? Isn't there a filter on the HPLC lines?"

I told him the vacuum filter was 0.2 um. He asked, "Can't you get an inline filter that's 0.2um? And if you buy HPLC grade, do you still have to filter?"

Does anyone have good responses for these questions? I've been preparing mobile phases for so long that vacuum filtering is just habit, so I didn't have any great answers for him.

Thanks!

EM

[mbulentcakar]

Hi,
Personally i don't filter HPLC grade solvents. But i remember many instances that 'hplc grade' water easily failed to hold the title (microbial growth). Also many buffers may have undissolved suspended particles which can easily clog the column inlet.
Surprisingly few systems have online 0.2um filters. I am a proponent of the idea to use the filters but still i would prefer that filter also does not get stucked. For this purpose i usually filter the mobile phase with a disposable filter so i have no concerns about clogging my inline filters, so no lost chromatography time.
I am aware that filtering also takes time but it's 'plannable' and it's conveniently easy to perform 'plannable' tasks in a laboratory rather than rushing for troubleshooting an hplc system for a clogging.
Best regards,
Bulent

[grzesiek]

I filter only mobile phases containing buffers, inline filters can get clogged and probably would pretty soon, I think you seen how does filter look like after filtering mobile phase, I wouldn't like that on the inline filter

[tom jupille]

"when to filter" has been a topic of contention for as long as I've been at this.

Your response to your colleague about the in-line filter was dead-on. The inline filters ("sinker frits") on most systems are 5 or 10 micron porosity. They serve to take out gross particulate contamination that gets in after mobile phase prep (and to keep the inlet line at the bottom of the reservoir). Something like a 0.5 micron will get fouled far to quickly to be usable at that location.

HPLC grade solvents are supposed to be filtered through 0.2 micron filters at the bottling point. Assuming a reasonably clean lab environment, the greatest source of particulates is probably salts and buffers. If you let HPLC grade water sit around open for a long time, then bug growth there can also be an issue, as pointed out by mbulentcakar.

My personal preference is not to filter pure organics at all, and to filter "mixed" mobile phases only after addition of salts or buffers. As they say in US car ads, "Your mileage may vary".

[skunked_once]

I always vacuum filter all of my mobile phases. Even HPLC grade solvents can become contaminated with dust or microbial fauna (as others have noted). The additional benefit of vacuum filtration is the degassing of the mobile phase. I have rarely needed to perform additional degassing in my applications.

[HW Mueller]

Are you talking about these 4cm+ filters when you mention "vacuum filtration"? I also used to do this regularily until I noticed that it was completely unnecessary to go through this hassle, in most cases, if you follow a few simple rules in handling high purity solvents. Thus, I would like to see a detailed protocol of how one gets rid of dust/dirt, and how one degasses solvents that way. To prevent more dust/dirt to get in then out due to filtration I had to use a considerable amount of the solution to wash the assembled filter and vessel. A meek degassing was obtained only if I hooked up the filter vessel directly to the HPLC, etc., etc.
I have used VALCO in-line sieves which lasted hundreds of liters before producing a notizable resistance (when that happened one could clearly see the dirt on disassembly).

[mbicking]

It's Friday afternoon, and I'm feeling a little bit argumentative, but isn't it about time that someone challenges the myth of vacuum degassing?

Let's look at this from a reasonably scientific standpoint. It is true that when we filter a solvent through a small pore filter, a vacuum will be required to create the flow. This low pressure will also cause outgassing of any dissolved air. The dissolved air was previously in equilibrium with the atmosphere (at atmospheric pressure), and now that the pressure has been reduced, the dissolved air re-establishes the equilibrium at reduced pressure (LeChatelier's Principle). And of course this is the situation as long as the (reduced pressure) is maintained.

So now that we are done filtering, we shut off the pump (there goes the source of vacuum), remove the filter (there goes the barrier to the outside world), and pour the filtrate into a mobile phase reservoir (which is, conveniently, filled with air at atmospheric pressure). So, exactly how long do we expect this solution to stay "degassed?" I'm guessing the equilibration time is a minute or so.

Does anyone have a O2 sensor? Let's do an experiment!

[Bruce Hamilton]

Myth, Myth! .

Anyone who has used a Waters 6000A solvent delivery system with AR grade solvents ( chromatography grades were only available from Waters at exorbitant premiums ) will attest to the value of vacuum degassing or helium sparging. The first two hours of the day could be spent priming the pumps if vacuum degassing wasn't used.

The idea was to premix solvents 90:10 for A, 10:90 for B, and vacuum filter through 0.45um. That would greatly eliminate bubbles caused by the lower solubility of gases in the blends, as there usually are more dissolved gases in the organic component than the water.

I preferred helium sparging, as it had less effect on the composition, and many of the system had aftermarket pressure relief values that kept a head of helium over the mobile phase - although Dionex had helium system inbuilt.

These days, vacuum filtration has just about gone the way of the dodo. I found alot of people broke the vacuum at the source, thus sucking back the debris into the solvent, unles they had a 0.45 um filter on the vacuum flask outlet. I've seen red rubber chunks from vacuum hoses in solvent reservoirs in some labs.

Milli-Q water systems supply on demand, and prefiltered HPLC grade solvents are relatively cheap. As with other users above, I may filter buffers prepared with salts, but even they are of much better quality, and a 47mm Durapore filter can last me months now, rather than days.

I degas solvent mixtures using an ultrasonic bath. I also have short lengths of 010"ID PTFE tubing as the sole vents on my sealed reservoir bottles, which hopefully also greatly decreases diffusion of air gases, thus reducing load on degasser module.

Please keep having fun,

Bruce Hamilton

[mbicking]

Bruce:
I'm glad someone would rise to the challenge.

I agree with He sparging. Have used it many times.

And I even have used a 6000A, although that was back when we had to fight off the dinosaurs to do any work. It was a nightmare to use because of the bubble problem, but I found it to be a problem with single solvents also.

If you (vacuum) seal your flasks, then I can see how they can maintain a degassed state, but if they are exposed to air, I don't see how they can stay degassed for any period of time.

Now, perhaps the problem is not the level of dissolved gas, but maybe the presence of microscopic bubbles, or some other physical phenomenon, which would explain why sonication also works.

But I would still like to see the O2 sensor experiment.

[HW Mueller]

Merlin, you don´t need an O2 sensor, a UV detector will do. Pass the differently airated solvents through without a column, whatch the baselines without resetting zero.
Now, I have seen a bothersome rise in baseline due to air diffusion into a completely stationary bottle from which the liquid was pumped, when this liquid had a high content of EtOH. I just saturated this mobile phase with air and worked with a fairly high, slightly more noisy baseline . . . was passable. No bubble (outgassing) problem.

[mbicking]

One of the other issues here is the Teflon tubing that is so popular to use between the mobile phase and degasser and pump. At the trace level, this tubing is very permeable to air. If you are using an electrochemical detector in a mode where oxygen reacts, all of this tubing has to be replaced by metal, even pieces just a few cm long. It's just another example of how hard it is to really eliminate these gases from a solution.

[Uwe Neue]

My degassing method has always been functioning and FAST: I prepare the (isocratic) mobile phase and put it under vacuum from the water aspirator. It bubbles vigorously for 15 seconds to half a minute, and then the bubbeling declines drastically, indicating that now the solvent wants to evaporate. So you are done.

Put it on the LC... no problems ever. Granted, today, the systems have more sophisticated ways of dealing with this on-line degassing, but my method was fast and did not change the solvent composition.

[Bruce Hamilton]

Hi Uwe,

You must be more careful than I - when I tried that technique some rather expensive mobile phases bumped so badly that large amounts went all the way to the aspirator, and I had to discard the remainder because of backflow contamination. Oops.. not enough patience...

I find pulsed ultrasonics ( on for 5-10 secs, off for 10 sec to allow bubbles to rise ) repeated twice works on a litre of mobile phase for me without vacuum.

If I apply a vacuum, then just 30 - 60 seconds continuous on is required - but you may need large volume of headspace, as plenty of fine bubble "foam" may quickly appear with vacuum application. As with straight vacuum technique, it's quite obvious when the dissolved gases have gone.

Please keep having fun,

Bruce Hamilton

[HW Mueller]

Circumstances at the university clinic (medical school) here in Giessen were such that I was practically limited to degassing via vacuum. I found it to be very reliable if 1 L brown screw cap bottles with the mobile phase, cap removed, of course, were placed inside a special dessicator and swirled during vacuum. These bottles were then directly hooked up to the HPLC very carefully.

[Victor]

There seems to be a lot of historical talk about degassing here...I used to do it this way or that way.. So I am not sure if people are still doing these things or whether they changed over to on-line degassers?

However, Bruce seems to imply that he first degasses his mobile phases by ultrasonication, and then fits them on to a system with an on-line degasser. Bruce-do you do this all the time, or only for critical applications like methanol-water with a low pressure mixing system?

I seem to remember an article by John Dolan which claimed that helium sparging was the most efficient way of degassing and I have no reason to doubt this. I never had any problem with degassing just using an Agilent on-line vacuum degasser, but I generally use ACN-water and a high pressure mixer, so this is hardly a critical application. Does anyone have more critical mixes where they have shown these degassers to be limited?

To go back to the original question, I tend to filter all aqueous buffers (but not HPLC grade organic solvents) through 0.22um filters. I note however that some sub-2um column manufacturers seem to suggest even filtering HPLC grade solvents. Does anyone have evidence that this is necessary? I am inclined to believe with others here that this could do more harm than good.