Optimization of amino column

[Smilla]

Hi!

I’m a student who is trying to optimize a reversed-phase Luna NH2 column (Phenomenex) (250x4.6 mm) (5µ). The compounds I’m going to separate are N-acetyl-D-glucosamine and D-glucosamine, and as the mobile phase I’m using acetonitrile and water with a flow rate of 1.0 mL/min and a column temperature of 40 oC. The detector is a RI-detector.

First I tested an 80:20 mobile phase (diluting the samples with water only) but then the N-acetyl-D-glucosamine came out in two separate peaks, and the D-glucosamine came out before, or merged with the solvent front.

Then I tested a 50:50 mobile phase (diluting the samples with the mobile phase, stock solutions diluted just with water though). Then the two peaks of N-acetyl-D-glucosamine had merged together a bit, but the peak of D-glucosamine still came out before, or merged with the solvent front.

I would like to have some advice from you experts, how I’m going to continue with my experiments. Shall I try to use a 40:60 mobile phase instead? Do I have to add any salt to the mobile phase, what salt and what concentration in that case?

Shall I dilute my stock solutions and samples with mobile phase or just water?

Does it play any role that my standard of D-glucosamine actually is
D-glucosamine hydrochloride?

Would be really glad and thankful if someone could help me with these questions.

Regards,
Smilla

[tom jupille]

Unfortunately, you have picked a difficult problem. Separations of sugars and amino sugars on amino bonded phase columns represent a sort of "twilight zone" between normal-phase and reversed-phase conditions (the name "HILIC", for "hydrophilic interaction chromatography" has been applied to this situation).

Basically, you need to run a series of experiments changing the ACN/water ratio in something like 10% increments, and plot retention as a function of %ACN. You may not find any usable region, in which case you need to switch to a different column.

[Smilla]

Unfortuntely I can´t switch to a different column... It´s my supervisor who has bought it, and I got the nice task to optimize it. And now I´m sitting here and getting more and more panic...

[MG]

In this separation, water is your strong solvent, and acetonitrile is your weak solvent (opposite of reverse phase). You may be seeing the split peaks due to too much water in your sample solvent.

[tom jupille]

Smilla, I understand your panic . . . however you're in a similar situation to swimmer caught in a rip tide: if you do panic and start thrashing, you will drown. If you take a deep breath, think, and swim parallel to the shore, you'll be fine.

That's why I suggested that you be systematic about things: run a series of experiments changing the ACN/water ratio in 10% increments, and plot the retention of your compounds as a function of %ACN. That will tell you whether you can get reasonable retention on that column or not.

If you can find a range of %ACN that gives reasonable retention, then you have a starting point for optimization.

And if you can't find a range that gives you reasonable retention, you will have hard evidence to discuss with your supervisor in formulating "Plan B".

[Smilla]

Can the separated peaks of N-acetyl-glucosamine be due to mutarotation or is it just because of too much water in the sample solvent?

Can anyone explain why mutarotation can be decreased by increasing the column temperature, or name a good reference that gives a good explanation of mutarotation?

[MG]

I don't know anything about mutarotation. But, since you are doing a HILIC separation, if your sample solvent contains more water than your mobile phase, it's likely to cause splitting or fronting of peaks. This is especially true if your separation is isocratic. One name for this phenomenon is "injection solvent mismatch". The same thing will happen in reverse phase if you inject a sample with too much organic solvent.

To avoid this problem, think about your mobile phase mixture. Ask yourself which solvent is the strong solvent, and which is the weak solvent? Then, make sure your sample solvent is either the same strength, or weaker than, your mobile phase. Also be aware that it can happen, in some cases, if the pH of your sample is too different from that of your mobile phase.

In gradient separations, you can get away with injection solvent mismatch if your injection volume is small enough.

I can't guarantee that by eliminating injection solvent mismatch, you will eliminate any fronting or peak-splitting problems. Other things can cause these symptoms. But you'll be eliminating one possible cause.

[HW Mueller]

You would probably need a catalyst like H+ to get mutarotation, anyway if it occurred on the column you should get only strong tailing, not two peaks. Most likely, if you can dissolve your compounds in ACN you wouldn´t get mutarotation but get MG´s effect instead.

[Smilla]

HW,

Dumb question maybe, but what is â€

[tom jupille]

Smilla: HW is referring to the previous post in which MG suggested that your peak splitting might be due to having your sample dissolved in too strong a solvent.

HW: "mutarotation" is the official term for interconverstion of the β- and α- anomers of a sugar like glucose in solution (remember, you came up with the Walton references a couple of weeks ago:
http://www.sepsci.com/chromforum/viewtopic.php?t=2793 )

I suspect that you're right about N-acetyl-glucosamine not interconverting (OK, mutarotating!) easily, and I'm not sure that the amino column can discriminate between the anomers in any case, so I'd be inclined to agree with you that MG's suggestion about diluent strength sounds more plausible.

[HW Mueller]

Thanks Tom for helping Smilla.

Smilla,
just tried to put a tiny sprinkling of humor into it and save on writing to boot.

[Smilla]

Ok, I get it... It really WAS a dumb question... But I´m Swedish and slow thinking...

I read an article (Journal of Chromatography A
Volume 1092, Issue 2 , 28 October 2005, Pages 246-249), where the authors suggested that monosaccharides could give partially resolved peaks due to mutarotation. That´s why I asked about it here.

[HW Mueller]

Well, if that´s the case (with swedes) then just look a little carefully and you will notice that all of us have a good bit of swedish in us.
For instance I had to look at my very old organic text (I wonder whether Roberts and Caserio is ubdated and still in use??) to remember that glucose mutarotates very slowly in aqu. solution at room temprature, and catalysis is via base as well as acid. Nevertheless, all suggestions you got should be rasonable.

[Smilla]

To all of you: Thank you very much for taking your time to help me.