Help - Gradient Peaks

[ejharrels]

Hello All,

I am currently developing a gradient method for a related compounds analysis. I am using an Agilent 1100 equipped with a DAD at 215 nm. My method utilizes a Waters XBridge C18 150 x 4.6 mm column with 3.5 um packing. The gradient is not linear but basically progresses as follows:

0 minutes - 70% K2HPO4 (pH 8.0) / 30% Organic (I am using both ACN and MeOH to achieve necessary selectivity)

28.5 minutes - 40% K2HPO4 Buffer (pH 8.0) / 60% Organic

I follow this with a 3 minute hold at 70% organic to rinse any "junk" off of the column and then reequilibrate at initial conditions for six minutes.

I have observed a gradient peak in blank and sample injections that elutes at ~ 22 minutes (~54% Organic) and interferes with a degradent of interest. After thoroughly washing the system and column, the peak is not present in early injections but consistently grows throughout a sequence and is much larger than my method LOQ by the last injection. In subsequent sequences in which no system wash was initially performed, the peak is present in early injections and continues to grow until the area stabilizes after several injections. The peak appears even if I run the gradient with no injection. I have tried multiple lots of water, K2HPO4, MeOH, and ACN. I have also switched instruments and columns. The peak is still present. The only thing that did help (somewhat) was switching from an HPLC grade ACN to a UV grade.

Based on these observations I do not feel the gradient peak is produced by my buffer. If it were, I would expect the peak to be large in my first few injections and then decrease in size as it is rinsed off of the system/column by my gradient. I am confident that the source of the peak is the ACN...this makes sense to me since I notice the peak increase in size as I continue to rinse with a high ACN content at the end of each injection (essentially building up more of the contaminent on the column). Am I misguided here?

If the ACN is the source are there any steps I can take to clean up the ACN before running my method? Would eliminating ACN from the column rinse at the end of each injection and using MeOH instead help?

Thanks...any input is greatly appreciated

[DR]

Search forum for "Empore extraction disks".
Poste therein describe how to clean up your water to eliminate this sort of peak (assuming that it is a function of water purity - which, despite your use of bottled LC grade water or a $$$ terminal purification unit, it usually is).

[ejharrels]

DR,

As indicated in my post the nature of the peak does not seem indicative of a contaminent from the water. I am familiar with empore disks but I am not certain that these will help with my problem.

[tom jupille]

I am not certain that these will help with my problem

But you won't know unless you try!

You *do* have 30% ACN in your starting mobile phase. If the contaminant is coming from the ACN, then pre-mix your A solvent and try the Empore trick.

[DSP007]

Hello all.

Firstly need to take into account that at 215 absorbs everything, even the oxygen of the air.
So the first thing you must show that the peak is not connected:
1) Bubble in the cell
2) reagents - all under suspicion, especially water and potassium phosphate.
3) In fact that is not "shit column", you seen.

2) What the concentration of buffer solution? Not ratio, mass concentration.
Further, there are no problems in mixing the solvent? Changing the refractive index of the mobile phase in the chromatogram can be reflected by false peaks, if the wavelength below 210 nm.

[carls]

After thoroughly washing the system and column, the peak is not present in early injections but consistently grows throughout a sequence and is much larger than my method LOQ by the last injection.

How do you "thorougly wash" your system? Perhaps the 70% organic at the end of the run is not sufficient to elute what ever is building up on your column. The contaminant does appear to be in your mobile phase but its not clear which component. Replace the filter/frit in the purge valve to eliminate this variable.

A good way to verify that the impurity is in your mobile phase is first thoroughly was both solvent reservoirs with HPLC grade water and methanol several times. Sonicate the the mobile phase filters in 50 v/v% methanol for 10 mins then 5min with water (for buffer filter) or methanol (for organic filter). Then make at least three gradient runs with your normal re-equilibration time then three runs with double the re-equlibration time follwed by three runs with triple the re-equlibration time. If the peak area is proportional to the eqilibration time the impurity is most likely in your mobile phase.

Then again, you might be able to sidestep this problem by stepping to higher %organic (85%) at the end of the run.

[ejharrels]

Thank you all for your replies.

I should have noted that I have performed the usual extened equilibration time test to confirm that one of the mobile phase components was the culprit...although I have confirmed that to be the case, the results don't provide much insight into which component is to blame as I start w/ 30% organic.

How do you "thorougly wash" your system?

The rinse procedure I referred to involved rinsing the sytem with water, nitric acid, more water, IPA, and more water. To rinse the column I followed Water's reccommended procedure. I then equilibrated the system and column with freshly prepared mobile phase and ran a sequence of 15 blank runs (no injection). For the first five injections my baseline was clean as a whistle. After that a small "blip" appeared at the usual time (~22 min) and by the end of the sequence had grown to ~35 mAU*s.

Replace the filter/frit in the purge valve to eliminate this variable.

I don't understand how this will indicate which component is to blame...can you provide further insight?

Last night a ran a sequence of 15 blank runs and modified the wash portion of my gradient to use 75% MeOH w/ no ACN and it seemed to help. I will repeat today extending the wash time a bit and increasing the MeOH to 85%. My buffer is 10 mM so I don't anticipate having an issue with salting out.

I have also requested a sample pack of Empore Disk Filters from 3M, when they get here I'll give them a shot...Any other advice is HUGELY appreciated.

[DSP007]

Well.
What qualifications acetonitrile and what is the optical density relative to the cell with the air?
Acetonitrile 240 from Panreak in 25 l metal drum produced a terrible impression of the presence of aromatic organics by GC-MS were able to identify benzene and xylene

[carls]

The filter/frit in the purge valve accumulates junk which can be the source of ghost peaks. By replacing it you can remove the filter as a variable.

When running the different equilibration times did the peak area return to the inital value when the equilibration time was returned to "normal"?

The fact that this impurity builds over time is unusual for mobile phase contamination when running an automated sequence of injections (i.e. precise timing between sucessive gradient runs). Normally the peak area of a mobile phase impurity will remain relatively constant from run to run even if it takes 5 runs to elute it.

Sounds like the concentration of the impurity is changing over time, i.e. impurity is coming from a component in your chromatographic system.

A few things to try:

Try bypassing the autosampler by connecting the pump outlet tubing directly to the column inlet and run the same sequence as before.

Try using water only (no buffer) and each organic solvent alone (no mixture).

[tom jupille]

I should have noted that I have performed the usual extened equilibration time test to confirm that one of the mobile phase components was the culprit...although I have confirmed that to be the case, the results don't provide much insight into which component is to blame as I start w/ 30% organic.

Okay, that narrows things down a bit. There are only a limited number of possibilities:
- the water
- the acetonitrile
- the buffer acid and base
- the mobile phase prep glassware (residual detergent).
- the pH electrode (if you put it into the bulk mobile phase).

- You have already exonerated the water and the KH2PO4 (did you use phosphoric acid to adjust pH? if so did you change that?).
- Changing acetonitrile lots makes *some* difference, so this is at least part of the problem.
If you were adjusting pH with the electrode in the bulk mobile phase, try another batch with no electrode/mobile phase contact (i.e., pull and aliquot, check the pH of that, then throw away the aliquot). I have seen contamination from pH electrodes.
- Hand-rinse new glassware and try making another batch of mobile phase.

If the organic solvent actually is the major source of the problem, try the Empore trick. That should take out most of the junk.

[DR]

I should have noted that I have performed the usual extened equilibration time test to confirm that one of the mobile phase components was the culprit...although I have confirmed that to be the case, the results don't provide much insight into which component is to blame as I start w/ 30% organic.

...
- You have already exonerated the water and the KH2PO4 (did you use phosphoric acid to adjust pH? if so did you change that?).
- Changing acetonitrile lots makes *some* difference, so this is at least part of the problem.
If you were adjusting pH with the electrode in the bulk mobile phase, try another batch with no electrode/mobile phase contact (i.e., pull and aliquot, check the pH of that, then throw away the aliquot). I have seen contamination from pH electrodes.
- Hand-rinse new glassware and try making another batch of mobile phase.

If the organic solvent actually is the major source of the problem, try the Empore trick. That should take out most of the junk.

OP's gradient is going from 30% organic to 60% organic and longer run times at 30% produce even larger mystery peaks.
This does pretty clearly implicate the water or additive other than the organic. Empore treatment of the water used to make A phase (best) or at least filtering the prepared A phase through an Empore should help.

Reason: The organic contamination coming in w/ the water just sits on the column until enough organic is present in the MP to bump it off. Running for longer times at initial conditions allows for concentration of more of the water's contaminant to build up on the column head. This is why longer equilibration leading to larger peaks is diagnostic of a water sourced impurity.

Tom, I think you may have crossed your NP/RP wires here... though I agree that the pH electrode is still a possibility.

[tom jupille]

DR, I see what you're saying, but I would argue that the information is consistent with contamination in the A solvent (not necessarily in the aqueous part of the A solvent). I can easily imagine a non-polar contaminant in the ACN which would be totally retained at 30% and elute at, say, 50%; in other words, I don't think we can exonerate the ACN.

Either way, though, the Empore filtration is woth a shot.

[ejharrels]

Thanks for all of the informative replies.

I have run the gradient using only water and organic (no buffer solution) and the peak was still present, eliminating the K2HPO4, H3PO4 (for pH adjustment), and pH electrode. The retention time was slightly different but I would attribute that to the varried pH (I should note the spectrum is identical to previous runs!)

The empore sample pack I requested should be here tomorrow. I will pre-mix my buffer and organic content to the starting conditions of the gradient and filter using a C18 disk. Based on the apparent non-polarity of the contaminant, it should be retained by C18 material under these conditions. I also intend to run another sequence only filtering the aqueous portion to eliminate that variable.

I am a bit uncertain as to how to treat my organic content though (for the rest of the gradient)...being that the contaminant is relatively non-polar I would expect it to stay in solution if filtering only ACN or MeOH with the C18 disk.

Any thoughts?

[tom jupille]

I am a bit uncertain as to how to treat my organic content though (for the rest of the gradient)...being that the contaminant is relatively non-polar I would expect it to stay in solution if filtering only ACN or MeOH with the C18 disk.

Generally, when you get "garbage" peaks from the "A" solvent, the problem is the accumulation of CRUD during the equilibration period before you inject and run the gradient. Typical equilibration is 10 column volumes. Once you start the gradient, the %B ramps up quickly enough that there's not much time for further accumulation to occur (at least, that's the fervent hope!). Even if the Empore filtration works, I expect you would still see some of the garbage, but at a much reduced level.

[DSP007]

Hi
D215 in organic solvent (out air D ) ? solvent D maximum (such as 260 nm) ?