Equilibration/resolution problems with new Acquity columns

[Nicki-Nitro]

Hi,

I am Project Chemist and develop UPLC methods for purity and assay for peptides with approx. 1000 - 6000 dalton on Waters Acquity UPLCs. One basic requirement for our methods is that the mobile phases must be suitable for MS. This means we are using mainly TFA containing eluents with 0.02 or 0.05% TFA in water and acetonitril (alternatively also formic acid, ammonium acetate or ammonium formate buffers). Since 1 year (we started more and more with method development on UPLC) we are confronted with a problem when using new Acquity columns for our analyses. After an equillibration procedure (see attachments) we do not obtain sufficient resolution with new columns. We see these problems with BEH PST columns, as well as with BEH Shield or BEH phenyl columns. Further we loose resolution, when we flush the columns afterwards with ACN/H2O 7:3 before storage. This leeds to a enormous efforts for several days to gain again resolution of the column.

We do not fully understand the problem. This problem is also present in other departments and at our customers, when the methods are transfered to them. One approach is, to flush the column with ACN/H2O 7:3 + 0.5% TFA at higher temperatures (60 - 80°C) for one day or over weekend. And we start to store our columns in ACN/H2O 7:3 + 0.05% TFA to "save" resolution (not to loose resolution again). We think that the problem might have something to do with ionparing/TFA/column/ligand/peptide.

Maybe someone of you has similar problems? I would like to know:

* Which equilibration procedure is adequate to gain resolution with new UPLC Acquity columns (using TFA-containing eluents afterwards)
* What happens exactly on the column with TFA?
* What is the secret behind ionparing/TFA/column/ligand/peptide?
* Why do we loose resolution after flushing the columns with ACN/H2O 7:3 before storage?

Here you can find more details (methods and overlay chromatograms). You can easily see my problems (not only including methods and pictures into this thread)

Equilibration procedure for new Acquity UPLC columns:
http://fotos.mtb-news.de/p/757789
Analytical method for one peptide:
http://fotos.mtb-news.de/p/757790
Overlay chromatograms of the peptide (stressed and unstressed) after the equilibration procedure:
http://fotos.mtb-news.de/p/757791
Overlay chromatograms of the peptide (stressed and unstressed) 2 days after:
http://fotos.mtb-news.de/p/757792
Method for the "TFA-flush":
http://fotos.mtb-news.de/p/757793
Overlay chromatograms of the peptide (stressed and unstressed) after the TFA-flush:
http://fotos.mtb-news.de/p/757794
Overlay chromatograms of the peptide (unstressed) after flushing the column for 30 min with ACN/H2O 7:3 and trials for reequilibration:
http://fotos.mtb-news.de/p/757788


I am looking forward to your feedback.

[danko]

I don’t remember exactly the pH tolerance of these columns but 0.5 % TFA sounds scary. And then 60 - 80°C. Why?
I think you kill your columns trying to save them.

Best Regards

[HW Mueller]

Checking all these chroms "hurts" so I am trying to get an understanding without them. After you "cooked" these columns for extensive time they worked again, so they are amazingly stable? If short treatment with that TFA solution deteriorated the column, but longer (harsher?) treatment improved them then you must have something in your samples that precipitates/adsorbes under some conditions, but is slowly removed under the "cooking" condition. (This could also be just a matter of equilibrium position). Did you run these columns and injecting only blanks (mobile phase)? Do they deteriorate?
I would approach this by substituting something for the TFA. (Those of you who have been around for some time will remember that I have had a huge share of problem with TFA).

[Uwe Neue]

A few answers:

1. If you do not get the separations that you want from a brand-new and well equilibrated column, you need to reconsider the chromatographic conditions. New columns are highly reproducible and guaranteed by Waters. If you need to treat the column with aggressive mobile phases to get the separations back that you had before, you must consider that the columns with your desired separations have been aged in your hands during method development, and that nobody will be able to guarantee you how you can get back to the same separations!!!!!!!!!!!!!

2. The BEH Shield column is based on a monofunctional silane and does not have the same low-pH stability as the BEH C18 column, which is based on a trifunctional silane with a special endcapping procedure. The BEH SHield column should not be used at a pH lower than 2, even at room temperature.

3. The BEH Phenyl column has a somewhat lower stability than the BEH C18 column. I am not able to judge all the details, but you will get better column stability at lower temperature.

4. To flush a column with 0.5% TFA at 60-80 degrees C over a day or a weekend is a guaranteed way to destroy even the best and most stable surface chemistry. Columns that have been treated this way should go straight into the trash can.

5. Your initial equilibration procedure is fine for a BEH C18 column.

6. TFA is a mildly hydrophobic ion-pair reagent. It sticks somewhat to the C18 surface. Its concentration can change the separation between specific peptides. TFA is also a strong acid that aggressively hydrolyses the C18, especially at high concentrations and with weak ligands.

7. For short-term storage, such as over a weekend, Waters recommend to keep the column in mobile phase. For longer storage, say a week or longer, Waters recommends to store the column in acetonitrile. This is a trade-off between hydrolysis adn speed of reequilibration.

Hope this helps!

[Nicki-Nitro]

@all:
Thanks a lot for your feedback. For the same question in the UPLC forum/community, I have not received any feedback so far. Maybe the forum is smaller or maybe the people there have not the same problems.

@danko:
You will find the answers to your questions at Uwe Neues feedback. He told us very clear about the stability of the different stationary phases of the Acquity columns. But this was already clear to me.

@ HW Mueller:
I also injected sample solution. When I injected 30 times sample solution, I saw the same effect. The peak separation became slowly better, but not as fast, as when I apply the aggressive TFA-flush. TFA is substituted by formic acid, when we obtain better resolution between the peaks. In most cases this is not the fact. Therefore we develop methods with TFA in the mobile phase. High baseline noise due to TFA is controllable, by reducing the TFA concentration (0.05 or 0.05%) or by adding more mixing chambers in line (three 50µL mixing chambers in serial). This decreases baseline noise and improves sensitivity. But this is not my problem.

@ Uwe Neue:
You mean, that I age the columns. That is likely partially the fact, depending of the stationary phase and the duration of the TFA-flush. I do not deny that. But when it is aging, that "improves" my separation, why does the peak separation disappear again, when I flush the column with ACN/H2O (without TFA) before storage? If the aged column shows the required peak separation, it should remain and not disappear.

@all:
I still think, that my problem has more to do with ion paring than with aging. As long as we do not flush/store the columns without TFA the separation remains. When we flush the column without TFA, the peak separation disappears. This is not the case in one method, this is the problem in many methods with TFA-containing mobile phases. I do not flush the columns with TFA because I it is funny and money pays no role. So far it was the only solution. If I have a better solution, I will change it immediately.

[Nicki-Nitro]

By the way, the retiontions times before and after the TFA-flush remains relatively stable. There is only a slight shift in the retention times noticably. For me the column resists the TFA-flush for some time. But I also admit, that we do not have long time experience with this procedure so far.

[Uwe Neue]

1. For the ACQUITY UPLC BEH RP18 column, you are clearly outside the stability range of the column, and your procedure kills the columns.

2. For an ACQUITY UPLC PST column, you should be within the stability range of the column, and you might encounter a slow equilibration procedure, although I have never seen it with TFA. As a reference, you might want to look at the following publication:
Slow equilibration of reversed-phase columns for the separation of ionized solutes Original Research Article
Journal of Chromatography A, Volume 1015, Issues 1-2, 10 October 2003, Pages 53-64
D. H. Marchand, L. A. Williams, J. W. Dolan, L. R. Snyder

3. The ACQUITY UPLC BEH Phenyl column does not undergo the same phenomenon as a C18 column as described by Snyder, so I am convinced that you are killing the columns in the same way as you are killing the RP18 column.

4. Let me repeat again:
For short-term storage, such as over a weekend, Waters recommends to keep the column in mobile phase. For longer storage, say a week or longer, Waters recommends to store the column in acetonitrile. This is a trade-off between hydrolysis and speed of reequilibration.

[Nicki-Nitro]

Thanks a lot for the publication. I have just ordered it.

For my knowledge, the BEH C18 column and the BEH PST 130 C18 are the same stationary phase. Waters declared that they produce a BEH 130 C18 material. Then they perform a test for the requirement for PST (Peptide Separation Technology). If the test is passed, the column becomes a BEH 130 PST C18 column. If the test fails, the column becomes a "normal" BEH C18 column. Let's say the PST column is quality 1a and the the BEH C18 column is quality 1b (or 2). Therefore the range for the pH-value is equal for both types (1-12).

[Nicki-Nitro]

@Uwe
By the way, you are not employee of Waters by chance?

[Uwe Neue]

Hi Nicki-Nitro,
Yes, I am working for Waters, and specifically in the department that develops stationary phases, columns and sample-prep devices. This is why I am able to give solid and detailed advice on such things.

[HW Mueller]

Now I am confused. When you run ACN/H2O through the column it gives bad results, if you inject samples it gives bad results, if you inject sample solvent (did you tell what it is?) bad results improve slowly, when you cook the column with TFA the situation improves very slowly (over night).

Did you ever test the columns with a standard testing mix after you cook them?

[Nicki-Nitro]

Dear Uwe,

are you from Waters/USA?

I am going to write you an email. I will send you my contact and if you agree we can discuss this topic offline. Concerning this problem, I have not get much support from Waters/CH so far.

Kind regards
Ronald

[Nicki-Nitro]

Now I am confused. When you run ACN/H2O through the column it gives bad results, if you inject samples it gives bad results, if you inject sample solvent (did you tell what it is?) bad results improve slowly, when you cook the column with TFA the situation improves very slowly (over night).

Did you ever test the columns with a standard testing mix after you cook them?

Yes you are right. This is indeed confusing and the problem I am confronted with. Actually not only me, but also other departments within my company. The sample solvent in this special case is 20% acetic acid. But in most cases it is mobile phase A.

We do not have a special test to test the columns and to compare quality, like Waters with the peptide mix for the PST columns. We only have a test and test mix for evaluation, if everything is proper installed (fittings, columns, etc.). But this test is not sensitive enough for this special problem.

[HW Mueller]

I am having trouble seeing any possibility for surmounting this problem without an independent (of your analysis) test of these columns at different stages of failure.
One thing still unclear: Does deterioration set in when washing with ACN/H2O without any analytes ever having seen the column before?

It would be nice if this did not go "under ground".

[Nicki-Nitro]

For me this special peptide is the right test substance. It is sensitive for this particular problem. When the resolution is fine with this method and peptide, the column works also fine with other peptides. But we do not see this problem with all our peptides.

Concerning your second point, before we flush the columns with ACN/H2O, sample solutions were always injected before. Otherwise it would not make sence, to clean/wash the column.

I can recommend the paper Uwe mentioned. This trace is "hot" and the problems described there are similar to our problems. I will follow that trace.