LC-MS Unresolved peaks: How to solve this problem?

[mobfesk2]

Hi Chromatographers,

I've been trying to separate a glycopeptide product from its starting material. I'm using a C8 column and detecting by LC-MS. My mobile phase is 0.1% TFA in water and 0.1% TFA in AcN. Temperature is 25C. My product has a tR of 12.29 mins and my starting material has a tR of 12.42 mins, which leads to 2 peaks that are not completely resolved. i.e. co-elution. Does any one know how I can get better separation? I'm using the following gradient:

Time (mins) %B
0.....................5
3.....................5
10...................15
15...................18
23...................30
27...................95
29...................95
30...................5
32...................5

I have even tried running from 15% - 18% B over 15 mins (10 - 25 mins) but still no fully resolved peaks

Thanks in advance for any suggestions

[Uwe Neue]

Flow rate? Column length and diameter?

[mobfesk2]

Flow rate? Column length and diameter?

Oh sorry... forgot to add this info. I'm using 1mL/min. Column length is 15cm and diameter is 4.6u

[tom jupille]

There are only six ways to move peaks around in reversed-phase:
- gradient steepness (you've already looked at that)
- temperature (try your same gradient but at a much higher temp and see what happens, if you see a change in selectivity then you can adjust from there)
- additive concentration/pH (try the same gradient with double or half the TFA and look for a change in selectivity).
- additive type (try formic acid instead of TFA)
- organic solvent type (not often done with peptides because of viscosity/efficiency issues, but might be worth try if all else fails)
- column type (a different brand of C8, or perhaps a cyano?)

Finally, you could try a longer (or smaller particle-size column) to give you more plates (i.e., narrower peaks).

You've already looked at gradient steepness, but from your post it wasn't clear if resolution improved. If it did, you might monkey with an isocratic hold (that's a long-shot, though).

Basically, line up a series of experiments exploring each variable in turn (in any order you want) and see which ones move your peaks around.

[mobfesk2]

Hi Tom,

Thanks for your quick reply! Much appreciated. When you say isocratic hold, how long should I hold at 15% for example? I'll try out your suggestions.

Thanks once again

There are only six ways to move peaks around in reversed-phase:
- gradient steepness (you've already looked at that)
- temperature (try your same gradient but at a much higher temp and see what happens, if you see a change in selectivity then you can adjust from there)
- additive concentration/pH (try the same gradient with double or half the TFA and look for a change in selectivity).
- additive type (try formic acid instead of TFA)
- organic solvent type (not often done with peptides because of viscosity/efficiency issues, but might be worth try if all else fails)
- column type (a different brand of C8, or perhaps a cyano?)

Finally, you could try a longer (or smaller particle-size column) to give you more plates (i.e., narrower peaks).

You've already looked at gradient steepness, but from your post it wasn't clear if resolution improved. If it did, you might monkey with an isocratic hold (that's a long-shot, though).

Basically, line up a series of experiments exploring each variable in turn (in any order you want) and see which ones move your peaks around.

[tom jupille]

When you say isocratic hold, how long should I hold at 15% for example?

I don't know (or even know that 15% is the best, for that matter!). It's a matter of "successive approximations" (sounds so much better than "trial and error", don't you think?). It might be worth a look if you resolution improved with a shallower gradient (isocratic is the ultimate in shallow gradients!), but it's a lot of work, which is why I called it a "long shot".

If it were my problem, I would try raising the temperature first (fairly quick to do, doesn't require preparing new mobile phase, and can be surprisingly powerful, particularly when combined with changes in gradient steepness). If that didn't work, I'd try varying the TFA concentration next.

The trick is to make a *big* change in whatever parameter you're looking at, and see if it has an effect on selectivity. If not, cut your losses and move to the next parameter. If there is a change in selectivity (i.e., if the peak spacing changes), then you can invest the time optimizing.

[Uwe Neue]

Did you get some improvement from the flatter gradient (15% to 18%B)? If yes, I would continue to work with this. For example, run the 15% to 18% separation again, but 1. double the flow rate, 2. change the temperature, 3. change from TFA to formic acid or change the TFA concentration and similar things.

[Keaka]

Try an extremely flat "slow as molasses" gradient (ie. 0 to 70-80%B over 35-60 minutes, or just flat out the middle portion of your gradient) and see what that gives you.

Peak shape won't look good but if you get better separation, you should be able to approximate more closely the %B that each compound needs for elution based on the flow rate, retention time and void volume.

You can shorten it up with isocratic holds etc. once you have that info.

For me personally, TFA + Mass specs should never be together. I'd give Formic Acid, Acetic acid or other buffers a try as well.

[Bryan Evans]

Is your C8 5um particle size? If so, I would decrease the particle
size down to 3um.

Also - a high bonding density ODS is generally recommended for
peptide analysis. There are exceptions - but I would always recommend
our C18 phase for peptides before trying a C8.