Chromatography Forum

Dear virtual colleagues,

I have been charged w/ obtaining the mass spectrum of a preNDA material my company manufactures.

A closely related compound uses a H20/THF gradient (both 0.1% TFA) ramping from 15 % B to 80 % B over 40 minutes at 0.6 mL/min with detection by UV. The parent compound elutes ca. 20 minutes.

My concern is that the high THF may significantly reduce the MS signal due to poor ionization of the analyte, or early anlayte ejection from the highly organic droplets

Is there a practical upper limit for the % organic? I'd like to have a good rationale to defend my position to managment.


Kind regards,
Robotjock

Actually, the reverse is happening: one does get a significant increase in sensitivity when the mobile phase has a high organic content compared to a low organic content. There are a lot of publications by Naidong Weng that advocate HILIC techniques over RP techniques due to the increased sensitivity. The usual expectation is a 5-fold sensitivity gain, but I have also seen extreme examples of a much higher gain in sensitivity.

I do not know if this has changed lately, but a lot of manufacturers used to recommend <30% of THF in the mobile phase when using ESI and especially APCI... For other organic solvents such as ACN and MEOH you can go up to 100%.

What Uwe said about the increase of sensitivity is correct, especially if you are operating at analytical flow rates. If the acidic additive is not important for the separation of your analytes, I would advice that you switch from TFA to formic acid...

Based on superior ionization in high organic mobile phases it appears that HILIC is the perfect chromatographic method for most small molecule LCMS applications, including compounds with moderate to high lipophilicity that elute from RP columns with more than 30% of water in the mobile phase. Does anybody know of a good source that compares all the available HILIC phases? I could spend a small fortune buying them all, but maybe I can narrow down my choices if I have good information about column chemistry and retention characteristics.

Pure HILIC is only good for very polar compounds. To get analytes with "moderate to high lipophilicity" retained in HILIC mode, you need to have a second retention mechanism. Such a retention mechanism is the interaction with silanols on silica HILIC columns, either by hydrogen bonding or ion-exchange.

I am from Waters, and we are selling the Atlantis silica HILIC column for such applications.

Uwe, thank you for your reply. The Atlantis silica column is certainly on my short list of phases I'd like to add to my column inventory for method development. I have noticed though that it was the least retentive of the four materials discussed in the Guo paper.

RP materials are easily comparable by looking at the carbon load. More carbon = more retention of nonpolar analytes. It would be very helpful if we had an equivalent method for HILIC columns that allows you to estimate the strength of polar retention.

For pure HILIC, silica is less retentive than other phases, for example amino phases. This is what the authors looked at in the Guo paper. On the other hand, if your analyte has basic functional groups, silica is MORE retentive than the other phases. Depending on the analyte, it could be MUCH MORE retentive.

If your analyte has basic functional groups, silica HILIC is your best choice.

A plain silica HILIC column will act as a cation exchanger when pH is raised above 4, as the -OH groups forms negative silanol groups on the surface. Basic compounds then bind more strongly and more buffer salts are needed for elution.

The ZIC®-HILIC column is based on a bonded phase having zwitterionic functional groups, while a silica HILIC column is just plain silica.

The ZIC®-HILIC zwitterionic groups maintain their charged but overall neutral hydrophilic form, independent of pH-value. This opens for a unique selectivity since the pH of the mobile phase can be used to moderate the dissociation and retention of the analyte(s).

The ZIC®-HILIC phase allows creation of a stable aqueous layer on the stationary phase, and holds a lot of water at the surface. This makes the column very robust in both isocratic and gradient mode, and it allows aqueous samples to be injected (unless too large volumes are used). The column also has a very good lifetime.

A silica HILIC column has all the problems associated with straight phase chromatography; adsorption effects, peak band-spreading (for some compunds), and slow equilibration. It is furthermore sensitive to changes in water ratio in mobile phase, making it difficult to run gradients due to slow kinetics.

More information can be found at www.sequant.com where you also can order, free of charge, the "Practical Guide to HILIC"

My suggestion is to add the zwitterionic ZIC®-HILIC column to your list.

/Patrik

Patrik, one more suggestion: How about doing the simple experiments, which I suggested previously, with your column, and reporting your results to us.

I have to disagree with Patrick. I have not seen any of the effects that he claims on HILIC applications. Slow equilibration with water is an effect in normal phase chromatography. In HILIC, you have a ton of water on the surface, equilibration is fast and gradients can be run without difficulties. There are commonly no peak shape problems when a highly pure silica is used, although I have seen such effects on low-purity silicas. Most importantly (as mentioned above) silica offers the opportunity to do ion-exchange on top of HILIC.

For a recent study on the water layer in silica HILIC, look at Journal of Chromatography A, 1192 (2008) 225–229.

For electrospray I've found an organic content of 60-70% optimum depending on the exact solvent, due to surface tension effects on droplet evaporation. With the system mentioned I would be careful of any PEEK tubing on the system which may swell in high %THF.

Would change the TFA over to formic or acetic if you can, if you're in positive electrospray you'll get a better signal, if you want to work in negative electrospray you'll actually get a signal.