Choosing a column for a polar analyte

A few years ago I was developing an LC-MS/MS method for psilocin in rat plasma at a preclinical CRO. I started where most people start, on a standard C18. The peak came off almost at the void volume. No retention, no separation from the matrix, and ion suppression at the same retention time.

I changed the gradient. Same result. I dropped the organic to 2 percent at the start, which is about as low as you can go before a standard C18 starts to dewet. Slightly better, still not a peak I could quantify.

The column was the problem. Psilocin is polar and basic, and a plain reversed-phase C18 has very little to retain it with.

Why polar compounds fail on a standard C18

Reversed phase retains by hydrophobic interaction. A polar analyte has almost no hydrophobic surface to offer, so there is little for the C18 chain to interact with. The compound stays in the mobile phase and elutes at the void.

Dropping the organic content is the obvious fix, and it only works down to a point. Below about 5 percent organic many standard C18 phases undergo phase collapse, also called dewetting. The bonded chains fold in on themselves. Retention drops instead of rising, and the effect is not always reversible. So the fix that looks free can damage the column.

Schematic. Dropping the organic raises retention on both phases until roughly 5 percent. Below that a standard C18 dewets and retention falls instead of rising, which is why the free fix stops working. Shapes are illustrative, not measured data.

What to buy instead

Four options, in the order I would consider them.

Option	What it does	Cost	Good for LC-MS?
Aqueous or polar-embedded C18	Runs in highly aqueous mobile phases without dewetting. The embedded polar group adds a second interaction that holds polar compounds	Smallest change from what you already run	Yes
HILIC	Retains polar compounds on a high-organic mobile phase	Slow equilibration. An aqueous injection distorts the peak, so the extract must be close to the starting mobile phase	Yes, and the high organic usually gives a real sensitivity gain
Ion pairing	Ion-pair reagent gives the analyte something to interact with	Suppresses ionization and contaminates a source for a long time. On a shared instrument it affects everyone else’s methods	Last resort
Derivatization	Adds a hydrophobic group so a reversed-phase column can retain the compound	Real sample prep work	Only when the other three fail

For an aqueous C18, ask for the phase to be described as “AQ”, “aqua”, or polar-embedded, and check the stated minimum organic.

Choosing between the first two depends on one question:

flowchart TD
    A["Polar analyte, no retention"] --> B{"What solvent is your extract<br/>actually in at injection?"}
    B -->|"Aqueous"| C["Aqueous or<br/>polar-embedded C18"]
    B -->|"High organic,<br/>or you can change it"| D{"Do you need the<br/>MS sensitivity gain?"}
    D -->|"Yes"| E["HILIC"]
    D -->|"No"| C
    C --> F{"Still no retention?"}
    E --> F
    F -->|"Yes"| G["Check the pKa against<br/>your mobile phase pH<br/>before changing column again"]
    G --> H["Ion pairing, then<br/>derivatization"]

What to check before you order

The minimum organic percentage the phase tolerates. If you plan to start at 2 percent, a standard C18 is the wrong purchase.
Whether the phase is described as aqueous compatible or polar embedded.
The pKa of your analyte and where your mobile phase pH sits relative to it. A basic compound run near its pKa gives broad and irreproducible peaks, and no column fixes that.
What solvent your extract will actually be in at injection. This decides whether HILIC is realistic.

The short version

If your compound is polar and your peak is at the void, another standard C18 will not help. Find out how polar it is, check the pKa, then choose between an aqueous C18 and HILIC based on what your sample is dissolved in.

One thing to rule out first. If the peak is small as well as unretained, the analyte may be getting lost during sample prep rather than on the column. Your recovery is low has the three injections that separate those two.

The columns I supply are on the HPLC and UHPLC columns page. If you tell me the analyte and the matrix, I will tell you which of these applies before you order anything.

If it is bigger than one question, and you want the whole method looked at rather than one symptom, that is the method audit. You send the conditions and a chromatogram, I send back a written diagnosis in 5 working days. CA$300, and if I cannot find a cause worth acting on you do not pay.