Decreasing retention time with increasing concentrations

[bisnettrj2]

Hello all:

I'm running a HILIC method on an Agilent 1100. I'm using a Phenomenex Kinetex 50x4.6mm 2.6-micron HILIC column with 86:14 A:B (A = 100% Acetonitrile with 0.1% TFA, B = 100% Millipore water with 0.1% TFA) at 1 mL/min and 30 degrees C, monitoring 257 nm, with a 10 uL injection volume.

Samples are dissolved in 9:1 Acetonitrile:Water with 0.1% TFA. My analytes are Paraquat and Diquat.

Diquat elutes first, and is non-target in this instance, but is included in my standard mix. Paraquat, my target, elutes second. 257 nm is the UV max for Paraquat.

My problem is this: My retention times are shifting earlier with increasing sample concentration. The first chromatogram below shows a 0.02 ug/mL standard overlaid with a 0.5 ug/mL standard. The second chromatogram shows all my standards overlaid, which range from 0.02 ug/mL to 1 ug/mL.

I am having a hard time accepting that I am overloading the column at these concentrations, but that's what it seems like is happening. Am I missing something here?

[Camisotro]

What is the retention mechanism for Paraquat and Diquat on this column? TFA is used most often as an ion-pair reagent and to make sure that everything is uniformly protonated. This does not seem to be a concern for the permanently charged Paraquat or Diquat. Have you tried changing the additive? I have not worked with HILIC columns but perhaps you might try a buffer instead of getting the stationary phase so highly protonated for your cationic analytes? I get the feeling that TFA could be contributing to poor retention here, which is why the column seems overloaded.

[HW Mueller]

This looks like a classical case of too much tailing.

[tom jupille]

To expand on HW's post, it's a textbook case of tailing caused by overloading.

[Flappytango]

Is there a particular reason you choose to use HILIC? There are other established method for these compounds. Aththough they seem to mostly ion pairing or specialty columns but they seem to give good peak shape. Restek, Agilent , Sielc, Suppelco and the EPA all have some sort of method for these analytes.

As far as HILIC you may be interested in this article, it includes your compounds on hilic..

http://chromatographyonline.findanalyti ... ail/706363

If you really want HIILC you micht try a column like the Waters BEH Amide. With its bonded phase you get a different mix of interactions and for certain analytes I have found this column to give much better peak shapes than the Kinetex column.

good luck

[bisnettrj2]

To expand on HW's post, it's a textbook case of tailing caused by overloading.

Even though I'm only loading 0.2 nanograms to 10 nanograms on column?

This is my first real foray into HILIC (where something actually retained; I did try a different analyte awhile back with no success in HILIC on this same column). Is this a common problem with quaternary amines in HILIC? With core-shell columns in HILIC? Or with quaternary amines in general? Or a problem of loading too much sample (10uL)?

At this point I'm not glued to this method. I do want to get away from the ion pairing method in EPA 549 as it's a bit 'messy' and cumbersome for me. I'm hoping to get something a bit simpler, and possibly mass spec compatible. Helpful nudges in the right direction are appreciated.

[Camisotro]

As far as HILIC you may be interested in this article, it includes your compounds on hilic..

http://chromatographyonline.findanalyti ... ail/706363

And notice that the method linked here uses ammonium formate @ pH 3.70 as the aqueous additive. Seems worth a try instead of TFA

[bisnettrj2]

Sure thing. I'll see if I can get some time to swap out my current mobile phase for a buffered one. Thanks for the tip.

*Edit - ctroster, I missed your original reply. I only had one notification of a reply, but obviously had more replies than that, and I overlooked yours the first time through.

[Camisotro]

Sure thing. I'll see if I can get some time to swap out my current mobile phase for a buffered one. Thanks for the tip.

*Edit - ctroster, I missed your original reply. I only had one notification of a reply, but obviously had more replies than that, and I overlooked yours the first time through.

No problem, let us know if it helps. This could also help with the matter of making it mass spec compatible, as you suggested you might want to do.

[tom jupille]

Even though I'm only loading 0.2 nanograms to 10 nanograms on column?

Unfortunately, yes. All that is necessary for "overload" tailing is that your sample is in a non-linear region of its distribution isotherm so that the top of the peak has a lower k' (i.e., moves faster) than the bottom of the peak. There's a "chalk talk" explanation on our web site: http://www.lcresources.com/resources/ex ... 070327.pdf

In principle, it's possible to have several non-linear regions in the isotherm, in which case you can overload at a low level, then peak shape improves (with a lower retention time!) at a higher level, and then tailing sets in again at an even higher level.

[Camisotro]

Unfortunately, yes. All that is necessary for "overload" tailing is that your sample is in a non-linear region of its distribution isotherm so that the top of the peak has a lower k' (i.e., moves faster) than the bottom of the peak. There's a "chalk talk" explanation on our web site: http://www.lcresources.com/resources/ex ... 070327.pdf

In principle, it's possible to have several non-linear regions in the isotherm, in which case you can overload at a low level, then peak shape improves (with a lower retention time!) at a higher level, and then tailing sets in again at an even higher level.

Thanks for the chalk talk! I've never thought of it that way. And presumably any change in additives or other relevant conditions can modify the shape of that isotherm as well.

[tom jupille]

And presumably any change in additives or other relevant conditions can modify the shape of that isotherm as well.

Exactly! That's why it's hard to generalize about specific causes and cures for peak shape problems.

[Vlad Orlovsky]

PQ and DQ in your case are retained by combination of HILIC and cation-exchange mechanism. With TFA in the mobil phase most of silanols are suppressed and you have very limited amount of them which are available for cation-exchange (you have some silanols which have lower pKa due to impurity inclusions, and the fact that PQ/DQ can shift equilibria a little). You can switch to a buffer with pH above 5 and this will increase capacity of the column and you should not see a shift in retention time with increased concentration. You will need higher buffer concentration to facilitate ion-exchange.
Alternatively you an try mixed-mode column which has 10 times higher capacity and will let you operate with less ACN in the mobile phase. You will achieve base-line separation with good peak shape:
http://www.sielc.com/Application-HPLC-S ... isc-R.html

[bisnettrj2]

Thanks Vlad. Today I prepared an Ammonium Formate buffer at 200 mM and pH ~3.8, and have premixed a 75:25 mobile phase of Acetonitrile:Am.Form. buffer for use with my samples, giving me 50mM buffer concentration in my mobile phase.

When I get back to the lab tomorrow, I'll flush out my system and column, equilibrate with my new mobile phase, and re-run using my new conditions.

If I get poor results with this setup, which mass-spec compatible buffer might you recommend above pH 5?

[bisnettrj2]

Update:

No work performed on this analysis this week. I had a rush project come in that absorbed much of my time; an unusual maintenance activity on another instrument required me to borrow the flow cell from this instrument; and finally I'm off tomorrow and this weekend for the US Memorial Day holiday, and to celebrate my wedding anniversary. In the meantime, I await a flow cell PM kit. I'll update with my new results once I get back up and running. Thanks again for the help and suggestions.