Downward Drift as %B Increases???

[AnalyticalWisco]

One more thought... if the buffer solution was pH adjusted with the wrong acid (e.g. Acetic), then the addition of the impure acid with higher absorbance may also result in the same downward drift as the ACN composition is increased. In the case of acetic acid, it is often of low purity and goes bad quickly resulting in an observable brown color.

I went back and reviewed chromatography for the last few months of runs. There's some variability in the magnitude of the downward drift, but it's always there. To me, this rules one-time errors (eg, incorrect reagent selection) which would have a more transient effect. That said, purity/contamination of the KH2PO4 and/or H3PO4 is definitely still on my list, as in my experience, these reagents can be in use for quite some time.

One other area I've been considering is the column being used (Waters Acquity BEH C18). I don't have a lot of experience with these 'Ethylene bridged hybrid' particles. I'm certain that the chromatographic conditions are well within the operating range of these columns, but I'm not sure if there are any special conditioning/equilibration procedures that should be followed when using them. For example, is there any reason that these hybrid particles might require longer equilibration periods, or a different post-run washing procedure?

Again, thanks for your input.

[AnalyticalWisco]

This might be a bit silly, but I only suggest it because I have done this before and achieved similar results. Check that your mobile phase lines are not backwards. i.e. A plugged into B at some point and vice versa...yeah, i did that.

Ha. I've done WAY dumber stuff than that... It's a persistent (ie, months-long) issue that's been repeatable across several systems, so I think we can run this one out.

Retention times, peak shapes, etc. are all right where we expect them to be, we just can figure this strange baseline profile out. Fundamentally, its not an issue (we sailed through validation without issue), but I'd feel a lot better if we understood why the baseline was dropping, even if we can't 'fix' it.

Thanks for responding!

[Hollow]

Hi

Just adding some more thoughts and tests:

- what about a contaminated solvent line/sinker frit?
-- then, same behaviour can be expected with other lots of water sources

- does it change when they switch to different solvent lines?
- are they using fresh and clean bottles?
-- do they wash the bottles in the dish washer? -> is the dish washer working properly? Any leftovers from the detergent/neutralizer?

- is the same issue seen on other systems too?
-- then the source seems to be the water/buffer components/glass ware

- what if they pass MP A through a C18 SPE or flash cartridge to trap any lipophilic contaminants?

***
and from my experience, the "normal" (non premier type) BEH columns are quite straight forward, about 5 column volumes (+vdwell) are sufficient for re-equilibration

also never had issues with broken degassers on Alliance & Acquity classic, at least not unrecognazable ones; the broken degasser pump was obvious and sensed by the system (probl. was due to dust and worn shaft bearings; 1x after about 7 years).

[Multidimensional]

"Multidimensional wrote:
(1) So never wait for an error, instead have the degasser profesionally serviced or replaced every 4-5 years of use)."

Reply:
"These systems are in a high-throughput cGMP lab, so they are serviced routinely (annual PM schedule, I believe). I've never actually looked into what is typically done during routine PM to verify Degasser performance. I'll have to look into that one, but I agree with you that the specific trend we're investigating doesn't really fit a degasser issue."

Comment:

--- PM service DOES NOT include any "real" service or cleaning of the HPLC vacuum degasser modules used in these system (to do so would cost the vendor as much as what they sell the PM service for and loose money). They simply put a "sticker" on them as "calibrated" or checked (basically, pure BS). As the degassers wear and/or when they break, (and all of the Alliance/Acquity degassers will be broken by the 5th year), they slowly introduce contamination into the system. *Waters quotes the replacement of the degasser module (Vac pump, tubing, chambers). The degassers have a limited lifetime and vendors choose to replace them,.not service them as this saves them the most time/money and insures the client gets a working degasser for the system. If you wait for any degasser errors to appear BEFORE servicing them, it is too late. The systems are designed to not show any errors until the system is destroyed, having already contaminated the vacuum and mobile phase flow path (*We have serviced thousands of these modules so see it first-hand). NEVER just replace a vacuum pump in a degasser as it will be contaminated and damaged after you re-install it back in the contaminated and damaged system. End users and/or the vendors do not have the specialized equipment needed on-site to test and repair these modules (That is why the whole module is replaced).
So many users are not aware of this and continue to "use" and leave contaminated vacuum degassers in-line placing their HPLC systems out-of-compliance.

[DR]

While I'm no fan of BEH columns, this does not appear to be a column issue to me.
Keep making new phases in "just rinsed" bottles using freshly opened bottles of stuff... You'll hit a cause eventually.
I also really like the "change the inlet frits" idea - many a PM goes by where that does not happen (if there's a drawer full of new frits somewhere in the lab, this is why).

RE: Waters degassers - on H class systems, there are individual cartridges for each line that share a pump. In theory, one can replace individual cartridges and save $. In practice, one cartridge fails and typically aerosolizes MP, contaminating the rest of the cartridges, or other cartridges fail in short order after one starts giving the user trouble. So, I suggest replacing all of them whenever one goes, or better yet - regularly as suggested by MD. Sadly, the degasser pump, which is the same as is used in Alliance, is available through Waters only and is not easily amenable to rebuilding, nor can it be sourced elsewhere on the cheap. They also last 6-10 years depending on use and there's no "servicing" them except to lube them if they get balky - this will get you through a run or two...

[TylerSmith123]

Hi Everyone,

My time to take a shot. Just kidding, however, I think a lot of people might be overlooking OP's comment about the repeatability of this issue: "
It's a persistent (ie, months-long) issue that's been repeatable across several systems, so I think we can run this one out"
If the issue is repeatable across several instruments/systems, then shouldn't we be looking away from the instruments and towards the sample? Unless, and I am not saying this isn't the case, all of the instruments in this lab have defective degassers/sinker/frits/lines. This would imply all of the instruments are running at roughly the same amount every day, running nearly-identical samples, etc. If this systemic issue was not addressed when it is happening to all of the instruments, I would be wildly surprised. However, it seems sample-specific rather than instrument-specific if this phenomenon can be repeated on other instruments in the lab.

[Multidimensional]

"Samples" do not change the gradient profile. We have seen entire labs filled with instruments, all setup incorrectly (very common) showing the same types of issues (so never assume it must not be the instrument related if "they all do it". These are training issues. As previously stated, incorrect mobile phase preparation and/or using the wrong instrument settings (composition, wavelength, etc) can cause the observed issue. Using a high absorbing additive, selection of the wrong wavelength/bandwidth values, OR turning 'ON' a software feature such as "Reference Wavelength' could be responsible, In any case, the troubleshooting needed to solve it should be straightforward.

[AnalyticalWisco]

If the issue is repeatable across several instruments/systems, then shouldn't we be looking away from the instruments and towards the sample? Unless, and I am not saying this isn't the case, all of the instruments in this lab have defective degassers/sinker/frits/lines. This would imply all of the instruments are running at roughly the same amount every day, running nearly-identical samples, etc. If this systemic issue was not addressed when it is happening to all of the instruments, I would be wildly surprised. However, it seems sample-specific rather than instrument-specific if this phenomenon can be repeated on other instruments in the lab.

Thanks for commenting. I appreciate you highlighting the persistence of this issue, as I certainly think does make many of the instrument errors/issues less likely, and points more in the direction of a systematic issue. Regarding the 'sample', I don't believe I highlighted this, but samples are prepared in 20mM phosphate, pH 3.0 in 20% ACN. This is stronger than the starting MP condition, but the injection volume is so small (1µL) that I wouldn't think this would make a difference. Note that we've also run true 'blank' injection (ie, zero vol) and we see the same profile. To me, this is a 'method specific' issue at this point. I do have some additional updates, but I'll provide those in a separate post...

Thanks again.

[AnalyticalWisco]

Update / New Information:
We received some additional information and new data this morning that I thought I would share. I've also got a new hypothesis that I'd like feedback on.

1) The lab confirmed that all recent runs have been on Aquity H-Class UPLCs with Quaternary pumps and TUV detectors. They confirmed that they have NOT been using a reference wavelength (so that hypothesis is officially shot down...)

2) They did perform a run on H-Class system with a PDA recently (again, no reference wavelength applied), and the profile did look significantly better (almost no downward drift); however, they confirmed that one other factor was changed at the same time for this run: They didn't filter the buffer for this run (they've typically filtered the stock buffer using a 0.2µm nylon membrane). So, this provides us 2 potential contributing (and currently confounded) factors for further evaluation: Detector type and filtration. I'm interested in anyone's thoughts on whether or a nylon filter may be contributing a contaminant to the MP.

3) I'm realizing that I did not specify the column temperature in my original post (50C). One of my colleague brought up the idea that changes in refractive index of the mobile phase may be contributing to the baseline drift. More specifically, if the mobile phase is at a high and/or inconsistent temperature when it reaches the flow cell, this can cause a higher degree of sensitivity to RI changes in the MP. We're thinking that this may align with the recent observation that baseline drift was different when using a PDA detector, as the PDA design may be less susceptible to this effect (maybe the PDA flow-cell has more thermal mass, thus stabilizing the MP temp in the detector?). Does anyone have experience with RI and/or MP temperature variation affecting the baseline?

[Hollow]

Thank you for the update.

I guess the filter may indeed be part of the issue. and/or the filter support.
Are they using sintered frits support? Maybe that is contaminated?

Maybe they can test and compare with a new filter device or clean the actual one with nitric acid (or similar) to get rid of any organic contaminants.

Maybe they can also try some other filters e.g. to hydrophilized-PTFE or regenerated cellulose?

[AnalyticalWisco]

I guess the filter may indeed be part of the issue. and/or the filter support. Are they using sintered frits support? Maybe that is contaminated?

Maybe they can test and compare with a new filter device or clean the actual one with nitric acid (or similar) to get rid of any organic contaminants.

Maybe they can also try some other filters e.g. to hydrophilized-PTFE or regenerated cellulose?

Thanks for the feedback. We've asked them to run using unfiltered MPs (same preps from the run on the PDA system) on the VWD system that they've been using. If the downward drift goes away, I think we'll be able to narrow the issue down to the MP prep. If we still see the downward drift, we'll focus on system differences.

I will note that it's only the stock buffer that gets filtered. Considering that equal amounts of that stock buffer go into the two MPs, I'm still not sure I understand how contamination of that stock would result in downward baseline drift, but who knows. I've been surprised before...

[Hollow]

How old is the stock buffer...? Has something grown in there?

Just to verify: is the chromatogram in the initial post a blank run? So all the peaks are contaminants?

Note, the system with the C18 column will act like a solid phase extraction cartridge, so also concentrating the contaminants from MP B, as long as the elution strenght is low enough.

MP A looks still as the main source to me.
Note the additional drop @27 min, when the MP changes back to A.
If not an effect of pressure or refraction etc, then this would fit with the SPE-model:
adsorbing until saturated, then the detector zeroes at a constant concentration, which then decreases as MP B increases, then flushing all the peaks out with high MP B and again.
(of course this would be a severe contamination if break through occurs, or something too big to get access to the stationary phase pores (SEC).)

How do they clean their glass ware? Do they rinse their bottles thoroughly before use?
(It may be, that there is something quite polar in the glasses, that will be more soluble in the MP A then in MP B. Then the same error would be seen on other LC systems as well, also with different lots of water)

[AnalyticalWisco]

How old is the stock buffer...? Has something grown in there?

They're giving the stock buffer a 14 day expiry. They've also prepped that stock at least 15 times over the past 4 months, so I don't think this is a stability issue with that buffer.

Just to verify: is the chromatogram in the initial post a blank run? So all the peaks are contaminants?

Yes. That's an injection of the diluent (which is 20mM buffer in 20% ACN). All the peaks you're seeing are 'gradient related'. I'm not sure I would put all those in the category of 'contaminatants', but yeah, there's a lot of 'stuff' being retained during the low-strength portion of the gradient which elutes during the 'wash' phase of the gradient. Luckily, all of our peaks of interest elute well before the wash. Would you/others consider the number/magnitude of peaks in the wash phase of that example 'atypical'?

Note the additional drop @27 min, when the MP changes back to A. If not an effect of pressure or refraction etc, then this would fit with the SPE-Model

I've always attributed that drop to the rapid (0.5 min) change from 100% MP B to 100% MP A, and the associated effects on RI/Pressure/Temp. The suggestion that this is actually a short window where contaminants are retained before they reach a saturation point seems plausible, I suppose. The model also fits the downward drift trend, assuming that MP A contains a higher level of the contamination than MP B. This model would seem to isolate the WATER as the source of the contamination, as that's the only component which is present at higher levels in MP A.

How do they clean their glass ware? Do they rinse their bottles thoroughly before use? (It may be, that there is something quite polar in the glasses, that will be more soluble in the MP A then in MP B. Then the same error would be seen on other LC systems as well, also with different lots of water)

I have no idea how they're washing glassware, but this is a commercial API manufacturing facility with a long history of GMP production/inspections/audits. I'm trying not to take anything for granted at this point, but glassware seems like an unlikely 'root cause' for such a persistent issue. Also, I don't think there's many potential contaminants that are going to be LESS soluble in MP B (which is a much stronger solvent, and still quite polar), but I could have that wrong.

Thank you very much for all the time you put into considering this issue, and for the constructive ideas. This is very helpful.

[DR]

::reread OP::

pH 3 with phosphate - not ideal, (you're not in a pH range where the PO4 provides any buffering capacity)... so even if your stocks are perfect and clean, it could still be the MP causing problems.

What MP glassware are you using? Schott glass can be problematic with ACN and very sensitive UV detectors...

[TylerSmith123]

Hi Wisco,

That is a very important distinction, and I'm sure many of us did not realize that was a blank/diluent injection. If this is a blank injection, there is certainly something on the column or in your injection matrix that is causing all of those peaks, and no, that amount of peaks is pretty irregular for blank/diluent injections, this should be a concern.

I also do not understand the logic here. If this is "stuff" being retained during the low-strength portion of the gradient, and supposedly elutes during the wash phase, then why is it still here? To me, it sounds to me like your methods have a washing portion that is not doing it's job. You will typically see changes in the UV as the gradient changes, but these two problems are pretty extreme in your users' case and certainly outside of the ordinary. These compounds may be eluting during the actual analysis as they are not being properly washed off and can contaminate your other peaks, so they are certainly contaminants. I'm surprised that at a GMP lab this would be ignored, but I've never been in one.