Sample Treatment for HILIC

[razorblademoon]

First of all: hello to everybody

I startet to work with HILIC and now I've got a problem. To get a good sensitivity, I need to avoid water in the injection. But: my analyte is just water-soluble. So if I concentrate my sample I'm not able to get my analyte into the acetonitrile without using a lot of water. Has anybody any experience with handling such samples? I would be very grateful for any ideas...

Thanks a lot in advance!

[Consumer Products Guy]

How about a small sample volume, under 3 ul?

[Mary Carson]

We've experienced same phenomenon. Some suggestions:

1. Get your analyte soluble with water, but add as much ACN as possible, preferably >50%.
2. Decrease injection volume. We usually use 5 uL, when our MP flowrate is 300 uL/min. On a RP, we could easily get away with a 20 or even 50 uL injections, or larger with a gradient, but we haven't had any success with larger injections on HILIC.
3. Go to very weak (95-100% ACN) starting conditions for your HILIC gradient. This may help off set the increased water in the sample.
4. Experiment with different HILIC columns. We've seen a wide variety in selectivity/ruggedness between the different manufacturers and even different lines within a manufacturer.
5. Vary the modifier. Change composition, pH, and ionic strength until you find conditions that allow the amount of water you need to solubilize your analyte.

[Bryan Evans]

In addition to the excellent advise already given, you may want to consider the particle size and column temperature.

The chromatogram below illustrates the difference between a conventional 5um NH2 column and Unison UK-Amino (3um)
for sugar alcohols done at 50C:

http://www.imtakt.com/TecInfo/TI313E.pdf

[razorblademoon]

Wow,
thanks for the fast answers.
1. First of all: I use an injection volume of 400uL for a 100x2.1mm column, with 5uL. I know it's quite a large volume. But there was no column overload yet and that's one easy way to decrease my limit of quantification. So I'd rather not decrease this.
2. I've already tried to begin with starting conditions uf to 97% ACN, but it didn't help.
3. I surely will try to vary with the ionic strength and the pH - so thanks for this hint!
4. We've got one other HILIC-column in the laboratory so I will try this one to. It's got even a smaller particle size!
5. The temperature is already optimized, but I'll try this parameter too.

Now I know a little more how to continue... thanks!

[Patrik Appelblad]

Your injection loop volume is 115 % of the total column volume, and it corresponds to a substantial extra column volume. You separation will be highly influenced despite you only inject 5 uL. I’d say it is the extra column effects that dramatically deteriorate your separation (due to dispersion). You should try to minimize it as much as possible.

In HILIC (as for RPLC), loadability will depend on system set-ups, sample solvents, etc, but if it is done carefully, there are good possibilities in injecting samples with high percentages of water. If you try to utilize maximum separation efficiency on the column, that is running at optimal flow rate, not injecting too large volumes or samples masses (that is prohibiting overload phenomenon), not using to large id tubing etc, you may very well inject samples with 50-70 % water in it. Also, as mentioned before, the ionic strength, type of organic modifier, and the pH will also be important issues.

If you then run a slow linear gradient starting at 95% acetonitrile, I would believe there is a good chance that you may improve your results.

[HW Mueller]

Is there anybody out there who would use plumbing which would send a 5µL injecion through a 400µL loop before going to the column?

[tom jupille]

Is there anybody out there who would use plumbing which would send a 5µL injecion through a 400µL loop before going to the column?

Probably not intentionally.

The "standard" plumbing configuration for a six-port injector reverses the flow direction through the loop when the valve is switched from "load" to "inject". That means that the injected sampe does not traverse the entire loop. When servicing a system, however, it is possible to cross-connect the pump and column lines (i.e., hook up the column where the pump should be and vice versa). In that case, the valve will function, but the flow will not be reversed, and the sample will traverse the entire loop.

That said, I wouldn't use that small a fraction of the loop volume, even if the valve were plumbed correctly.