Regarding pH, this partly depends on which Poroshell 120 HILIC resin you're using - the HILIC-z can go up to pH 12 while the other two top out around pH 7.

Otherwise, retention time in HILIC increases with compound ionization, so you will resolve this metabolite from less polar background better at a pH where it is ionized (see if you can find pKa value(s)). If it is a complex/diverse mixture of analytes, I'd use something between pH 5-7. 10 mM ammonium acetate with 0.02% acetic acid might work (~pH 5). If you need to go more acidic, you can try 10 mM ammonium formate with 0.2% formic acid (pH3).

HILIC is a mode of chromatography. It may or may not be a good choice to start with. However, screening for a suitable analysis method is usually performed using multiple methods and often multiple column types too. The closest thing to a "generic" gradient (we discourage any generic methods as a 'method' is suppose to be selective for the sample under analysis) would be NOT one method, but many. No single method is ever used for screening and would not cover enough options or conditions to know if it should be pursued (you would misinterpret many results). *ACN/Water (with or without any buffer) by itself may be a poor choice.

For your RP application, you may want to set up an automated screening system using gradients with (a) ACN/aqueous; (b) MeOH/aqueous; (c) ACN/aqueous acidified; (d) ACN/aqueous; with base; (e) MeOH/aqueous acidified; (f) MeOH/aqueous with base. Those options would provide for a neutral, acidified and slightly basic option and improve solubility overall. To reduce the total number of permutations, you could start with just the neutral or acidified versions. Note: To prevent problems later on, please not use TFA in any of the methods as the "acid". Stick with formic acid since LC-MS is applicable

William Letter Thank you for your answer! The HILIC column is not to start with, as I've already purified the metabolite to an active peak on the c18, but a couple of other polar compounds are proving difficult to separate from. I was hoping to design a generic HILIC gradient to see if I can gain better separation, given that I do not know the metabolite of interest's structure, and high-res mass spec suggests that it is a new compound. Your answer does give me the idea to take the active peak I've separated on the C18, and see if re-running it at a different pH will pull it away from the other molecules in the active peak.

BTW: No "HILIC Columns". Some companies put labels on them which imply this for mostly marketing reasons.

You need only apply good HPLC method development principles to find the answer. pH is just one of many parameters which must be investigated. Learning HPLC method development takes many, many years to learn just the basics of so try and find a professional chromatographer to help you with the project so you collect valid data.

Please do not get stuck on a specific 'mode' of chromatography, though HILIC can be useful for those compounds that are hard to retain on a typical C18 column with lots of aqueous phase. HILIC is actually not used that much, but has become very popular with sales people in the last few years. They will even have you believe that they sell actual "HILIC Columns", which is just a made-up marketing tool. Using the chromatography mode, HILIC, involves a mobile phase change only, and almost any type of column can be employed (silica, amino, propyl, C8, C18...), as applicable, using routine HPLC method development. A far more common solution is to use one of the super-high carbon loaded (C30) columns with conventional mobile phases to retain those peaks, if that is what you in fact need? Please seek the help of a professional chromatographer that is local to you for assistance with your project. These web pages are not suited to teaching you basic chromatography method development principles (which will take years to learn).

Hydrophilic Interaction Chromatography (HILIC); https://hplctips.blogspot.com/2012/03/hydrophilic-interaction-chromatography.html

10 mM ammonium acetate pH'd with acetic acid is a pretty common HILIC buffer system for LC/MS. You've mentioned a loss in sensitivity (assuming that's what you meant by the MS being worse) - you might just try negative mode on the MS just to see if you get better signal (go up to pH 6.5 buffer with negative mode, which is about the highest your column will handle). The change in sensitivity may be an indication that you need a more acidic buffer like the 10 mM ammonium formate with formic acid mentioned earlier. Again, without pKa info, you will need to do this through trial and error. If you are preparing 10 mM ammonium acetate, you shouldn't need much acid to bring the pH down, though more will be needed in the acetonitrile phase. Also keep in mind that your column may become charged as you go from low pH 3 to higher pH's, which can help or hurt depending on the unknown metabolite's chemistry. Also, make sure you always have a little water in the mobile phase (not 100% ACN) and that you condition the column well (see datasheet) before injecting. Best to try a few pH conditions in positive and negative mode and see what works best.

Dilute your analyte with buffer B ( high percentage ACN ingredient) this is essential to get a gaussian and high signal at HILIC. Your injection composition must be weaker than the initial organic composition of the gradient. Similar to RP but the reverse of the condition. On the other hand, adjust the best mobile phase condition by experimental query, in this case, I suggest increasing your volatile buffer concentration 20mM, 50 mM, 100mM, etc. It can affect ionization. Additionally, dilute and set the buffer strength of the injection composition as the same or very similar to the initial buffer strength of the gradient. Furthermore adjusting the desolvation conditions are also critical at HILIC-MS applications. Low capillary (ESI needle voltages) and well-optimized desolvation flow and temperature are needed.

good luck, Emir

İsmail Emir Akyildiz Thank you very much for your answer! It would appear that my issue may be due to the injection composition. If at the start of my gradient the solvent levels are 95% ACN and 5% aqueous buffer, could I use 75% ACN 25% methanol as my injection composition?

Timothy, your injection composition is not ideal given your starting conditions - organic strength is lower and you have no water which will mess up the hydration layer on the HILIC resin. I'd try using 95% ACN at the same pH you've selected, but being that this is a polar molecule, you have to wonder if you'll have solubility issues at 95% ACN (90% ACN for injection and starting conditions better perhaps?). If it separated using reversed-phase, why are you switching to HILIC anyways?

Andrew K Ottens Thank you, I will try injecting within 90% ACN and adjust the gradient accordingly so that it starts at 90%. I am using HILIC to further isolate the metabolite of interest after I've already done my best to purify using C18. With C18 I am able to isolate a peak containing the metabolite of interest, but it still contains enough other compounds that would cause issues with NMR. My current workflow is crude extraction with ethyl acetate--> evaporation and resuspension in MeOH (ethyl acetate caused issue with clogging in the HPLC) --> preperatory-hplc with c18 isolating the active peak w/ a time based fractionation collection --> evaporation and resuspension (previously used methanol here but will try 90% ACN) --> HILIC. The goal is to further isolate the metabolite of interest from the other compounds present in the peak so that I can proceed to NMR. If you have any suggestions for alternative routes besides HILIC I would be grateful. The compound appears to be polar/amphipathic and has a MW of

Generally speaking, RPLC would be used for the final polishing, not ahead of HILIC. You might be better off trying the HILIC first (or another preparative phase like ion exchange) and then RPLC.