Loss of retention on a HILIC

[mah]

Hi all,

I've tried to develop a LC-MS/MS method for bases and zwitter ions on a Atlantis HILIC column (2.1x100, 3µm + 10 mm guard) with a Agilent 1100 series system coupled to a MS/MS (Sciex API3000).

Mobile phase A: 60/40/0.1 methanol/water/TFA (resulting in pH 2)
Mobile phase B: Acetonitrile
The flow rate is 250 µL/min and injection volume is 10 µL, running gradient: 0-1 min 5%A -> 42 min 20%A (tG~0.4%A/min).
Vial solvent: 75% acetonitrile, 25% methanol.

Problem: a total loss of retention after 5 replicated standard solution injections

PS: I notified that the column pressure is rather low, approx. 1500 psi.

/Martin

[Kazimierz]

Dear Martin,

Problem is probably focused on changing pH because of gradiend.

1) Please try regenerate column (acid in H2O+small acetonitril overnight recirculation), next H2O+small acetonitril, next acetonitril many hrs.

2) I suggest construct isocratic gradient and both solvent A and B with pH=2.
If possible eliminate methanol. Recirculate with phase many hrs (even overnight).

3) Injection solvent - only acetonitril (or even 90% acetonitril in H2O)

[mah]

Dear Kazimierz,

Thanks ! - I'll try the regeneration this weekend...

Do you think, that an addition of ammonia formiate (5 mM) to the mobil phase would do any good ?

best regards
/Martin

[Kazimierz]

Yes, but I suggest initially try with acid only - for the check if column is good.

[Einar Ponten]

The description of your conditions were very detailed and that is indeed helpful and nice.

Still, there is probably not a straight answer. You inject bases and the injection volume is about 3% of column volume. As I understand it you don't pH adjust your samples and therefore the bases are allowed to raise the pH, which will turn the silica material into an cation exchanger due to silanolic interaction at neutral pH. Thus, injected analytes are highly retained.

In the first place I suggest that you add buffer to the sample (or dissolve in mobile phase). Decrease injection volume to 2 uL..

If this doesn't work I suggest that you try alternative columns and I am more than happy to give you suggestions....

[Uwe Neue]

I do not yet understand where the problem is coming from.

Have you had previously good results with this method, or is this the first time you tried?

I do not believe that you loose retention due to loss of water on the column, unless there is a reequilibration step or column wash step in your method.

It could very well be that TFA is too strong an acid and deactivates the column, i.e. reduces the retention of the bases. Did you see a decrease from run to run, or was the decrease sudden?

I hope we will find a clue what went wrong from the answers to these questions.

[mah]

Dear Uwe,

This is the first time I tried the method. Between each run there is a reequilibration step of approx. 20 minutes (5%A).

The decrease in retention where sudden, i.e. no retention in the following 5 injections.

Maybe it is a good idea to try citric acid (pKa 3.1) to reach pH 2 instead of trifluoroacetic acid ?

/Martin

[Uwe Neue]

The sudden change in retention together with our inability to figure out, what the problem is, indicates that the problem may not be associated with the HILIC technique. Please investigate other problems such as pump failures etc.!

[mah]

Hi again,

After regeneration of the HILIC column (as Kazimierz described), I have equilibrated the column in 100%B for approx. 12 hours;
A: 95% [0.1%TFA in H2O (pH 2)] / 5% [ACN]
B: 5% [0.1%TFA in H2O (pH 2)] / 95% [ACN]

and done some injections of standards in
a) 100% ACN
b) 95% ACN and 5% [0.1% TFA (pH 2)]

The injections resulted in no retention !?
-thereafter I tried the Test Chromatogram of the column (retention of cytosine). The column passed the test - the test was okay !


/Martin

[MG]

Perhaps your analytes have no retention in HILIC under the conditions you are using. Perhaps in your initial analysis where you observed retention, your column was not equilibrated or something else was wrong.

[Kazimierz]

Dear Martin,

I understand, that you are sure that unretained peak is your analyte (MS/MS detection).

If so, please manipulate with pH. Also I found literature with phase:
acetonitrile-H2O-acetic acid-trifluoro acetic acid (93:7:1:0.025) for HILIC silica.

If possible please include also uracil and quanosine for the test of your column.

[smiley]

about the test chromatogram:
never believe in that, my column lost resolution totally on my compounds, and still it passed the test!
i believe what MG said is probably right,
try luna silica column to compare with the atlantis one, my own experience is the luna is better than waters.(no offense to Uwe and waters)

[Uwe Neue]

Smiley,

In the context of this problem, your statement is simply nonsense.

[smiley]

well, not really,
what i tried to say is at least you get another similar silica based HILIC mode to test if it is anything related to your waters column, and if it isn't, it will prove that the atlantis is good in HILIC mode and has nothing to do with the problem.
don't you think?

[Kazimierz]

If realy these basic compouds are not retained on this column it is other possibility with Flanagan method published in J Chromatogr. 1980's

1. Componds should be basic which not eluted with methanol on silica.
2. Mobile phase: ca 0.05% perchloric acid in methanol (possible modification
with organic solvents or pH or salts - retention pKa dependent).
3. Method is good even for lipophilic drugs like amiodarone

http://www.ncbi.nlm.nih.gov/entrez/quer ... ds=2987280