Chromatography Forum

Hi everyone

I'm analyzing simple sugar standards for the past 3-4 weeks now and I am getting exactly the same retention time no matter what sugar I inject into the column. My supervisors have not heard about this problem before and we have tried everything to try and fix it but the problem still doesn't go away.

We are using an Amino NH2 column with a RI detector coupled with an ELSD detector, with mobile phase 80:20 ACN/water up to 90:10. I have tried both isocratic flow and gradient and showed no improvement. I even tried using millipore water and showed no change. We played with the flow rate, and the column incubator but no good. We are thinking of buying a carbohydrate column but it is costly and we don't want to risk buying something expensive and then finding out it wasn't the columns problem. We are now on the verge of giving up and changing to a GC method.

Does anyone have the same problem as I do? If so, what could you suggest on doing?

Thanks very much
Wilfred
Auckland, NZ

Does the peak appear when you run a blank sample? How old is your column? Some of those Amino NH2 columns can be a bit iffy when exposed to too much water. Have a look online, your column supplier may have information regarding care and maintenance or even have an application note available for your analysis.

Can you be more specific in some of the technical aspects that you are describing:

1) What is the size of your columns and how much are you injecting?
2) When you perform gradient elution what is the composition of your mobile phase A and B?
3) When you say you even tried millipore water, what were you using before?
4) What do you expect to see with your RI detector when performing gradient elution?
5) Did the method work before or are you trying to set this up?

Following up on Kostas reply please also let us know the retention time that always is the same.
It sounds like your sugars are not being retained and emerging in the void volume.
What sugars have you tried?

also, what sugars are you trying to separate, and from where did you get the method? Using a Phenomenex Luna-NH2 column I find that separation is largely down to size, and separation of two very similarly-sized sugars is poor. Also, Phenomenex claim that pH and buffering are very important to good separation, but I don't know if this is specific to their NH2 column or a general feature of sugar separation in Hilic mode on amino columns.

I'd also point out that gradient chromatography isn't recommended with IR detection.

Waters have some ELSD based applications here http://www.waters.com/webassets/cms/lib ... 3438en.pdf

Thanks everyone for your reply, much appreciated. I will try answer all your questions as well as I can.

Richiekichi: No peaks were found when I run blank samples. I have used two amino columns so far. One of them were from our previous student. We have flushed it with 100% ACN for an hour prior to use. It wasn't working, so we used a brand new one of the same model. Flushed it with isopropanol before use but still isn't working. We tried ringing Phenomenex but we already did what they adviced us to do.

Kostas: Column size is 250 x 4.6mm (5um). We are injecting 20uL at a time. Composition of the mobile phase A is simply 100% ACN and phase B is 100% millipore water. We start off running at 80%A:20%B and increase to 90%A:10%B around 10 minutes. Before millipore water, we used deionized water which I found out it was very unstable and the blank runs wasn't straight. I do not know what to expect when using gradient with the RI, but my supervisor told me to give it a go and see. I am trying to replicate the method supplied to us from Phenomenex themselves. They successfully seperated sugars with the same column we are using, same mobile phase (isocratic), at 3mL/min, column temp at 40 degrees with a RI detector. I did all those things but has not worked. All sugars are eluting out at the same time.

Sepscientologist: The retention time changes a little whenever I change the composition of the mobile phase. Today I tried (again) at 80%A:20%B and the sugars came out at around 5.5 minutes. I even injected a sample of caffeine and it came out at that time! When I tried 85%A/15%B, the sugars came out at around 7.5 minutes. I guess you are right, none of the sugars or whatever I inject is not being retained. I have tried: glucose, sucrose, mannose, xylose, fucose, fructose, arabinose, rhamnose, and galactose.

lmh: We are trying to seperate the above sugars. I am trying to analyze the sugar composition in seaweed and that will be another big hurdle for me since I cannot even get the standards running! The main method we are using is from Phenomenex, while a few others were from articles but they are also very similar. I will keep the pH and buffering capacity in mind. thanks.

JGK: I had a look at their method on sugar seperation and looks very different to the method Phenomenex supplied to us. Thanks. I might give that a go.


Thank you all for giving your advice ~ Much appreciated!

Wilfred

you have to change your thinking when working with NH2 columns as these work in HILIC mode. Therefore the strong eluent is water and not ACN!

So in fact, your gradient goes in the wrong direction (strong > weak)
Nevertheless RI detection is not compatible with gradient elution because the reference cell stais with the initial solvent and the change in the refraction of the eluent during the gradient is possible more than the change when your compound elutes.
You have to stick with isocratic mode with such a detector.

The way to go now, would be to change your % of water to less until you get enough retention and possible separation starts. Continue your serie with 90/10 and 95/5 and see what happens. Make sure to extend your acquisition time as well.
If you still got not enough retention, try to exchange or dilute the water part with methanol and subsecquently with isopropanol (e.g. 95%ACN / 2% H2O / 3% MeOH or and so on)

If your sample is dissolved in water, try to inject less (maybe more concentrated) or try to exchange/dilute the water with MeOH.
(it's like you would inject your sample in 100% ACN on an 20%ACN initial eluent on a C18).

Thank you so much Hollow. I will actually give that a try also ~ I have already tried 90:10 ACN/Water, so will try 95%:5%. If not I will mix in the methanol and see.

Thanks again

Hand mix your 80/20 ACN/H2O mobile phase and use one channel of the HPLC pump. Be sure to purge the HPLC system with mobile phase. The system purge volume depends on the system but an Agilent 1100 with the stock degasser, for example, requires ~ 20mL.

Your sample/injection solvent should ideally be mobile phase but use at least 50/50 ACN/H2O.

also make sure to properly equilibrate your column. HILIC possibly needs more volume to re-equilibrate, try also 10 (or even 20 column volumes) of mobile phase (with your column this is about 25ml or more)

Hand mix your 80/20 ACN/H2O mobile phase and use one channel of the HPLC pump. Be sure to purge the HPLC system with mobile phase. The system purge volume depends on the system but an Agilent 1100 with the stock degasser, for example, requires ~ 20mL.

Your sample/injection solvent should ideally be mobile phase but use at least 50/50 ACN/H2O.

By purge your "system" I mean be sure mobile phase has purged all flow paths in the autosampler including the needle/loop in addition to purging the tubing, degasser, pump heads, etc.

Also, If your system (which HPLC system are you using?) uses a needle flush/wash use mobile phase as the (weak) wash.

The void volume of this column is ~ 2.9mL so you can use this to calculate the void time at each flow rate.

Sugars are commonly analyzed using ACN/H2O (no buffer) mobile phases when using amino columns. Buffers are highly recommended for HILIC separations when running ionizable compounds.

Carls is right and Hollow too. Can I add, injection solvent! If you are dissolving your sugar standards in water an injecting them, they will all come through in the void volume. Honestly, I've been there and done that. Water is such a strong eluter in Hilic mode that nothing will bind if injected in water. The usual problem I've had in running hilic is finding an injection solvent that is sufficiently aqueous to dissolve the polar samples, but sufficiently organic not to elute them straight away. I would not recommend injecting in anything more aqueous than 50% acetonitrile, and even then, keep the injection volumes tiny. In general, I try to get to 90% acetonitrile.

These will in any case be very hard to separate.

Carls is right and Hollow too. Can I add, injection solvent! If you are dissolving your sugar standards in water an injecting them, they will all come through in the void volume. Honestly, I've been there and done that. Water is such a strong eluter in Hilic mode that nothing will bind if injected in water. The usual problem I've had in running hilic is finding an injection solvent that is sufficiently aqueous to dissolve the polar samples, but sufficiently organic not to elute them straight away. I would not recommend injecting in anything more aqueous than 50% acetonitrile, and even then, keep the injection volumes tiny. In general, I try to get to 90% acetonitrile.

These will in any case be very hard to separate.

Just to back up the excellent answer from Imh with a good paper:

Journal of Chromatography A
Volume 1217, Issue 52, 24 December 2010, Pages 8230-8240

A systematic investigation of the effect of sample diluent on peak shape in hydrophilic interaction liquid chromatography
Josephine Ruta, Serge Rudaz, David V. McCalley, Jean-Luc Veuthey and Davy Guillarme

Abstract
The aim of this study was to evaluate the importance of sample diluents to improve peak shapes in hydrophilic interaction liquid chromatography (HILIC), using low molecular weight (<1000 Da) analytes as well as peptides (with MW ranging between 1000 and 6000 Da) as model compounds. Various solvents were tested including water, acetonitrile, methanol, ethanol, propan-2-ol, dimethyl sulfoxide, and a number of combinations of them. For the analysis of small MW compounds, best peak shapes were obtained with sample dissolved in pure ACN but, IPA or a mixture of ACN/IPA (50:50, v/v) could represent a viable alternative in the case of solubility issues with pure ACN. For drug discovery applications, DMSO can be employed but in combination with at least 80% of ACN. For peptides analysis, acetonitrile, EtOH and IPA as sample diluents, provided similar chromatographic profiles, but pure EtOH or IPA were recommended to limit denaturation and samples solubility issues. Finally, whatever the nature of the compounds, it is recommended to add the lowest amount of water to the sample diluent, to maintain suitable peak shapes.

Best regards

Having been where you are coming, from it looks exactly like the sample solvent is the problem. When I am working with botanical extractions I dissolve in pure water and then bring my solutions to volume with ACN. This works rather well however I do have to be careful with my injection volumes. When I need a -NH2 column for sugar separation I usually go half the volume I do with strictly a vendor carbohydrate column or the water in the solvent eliminates all interaction with the sample and the column and I get one peak.
I would also agree with the ratio of 50/50 ACN/DI, you will find this will help crash out any proteins that may be dragged along through your extraction and keep your column alive a little longer. Go any higher on the organic and I start to precipitate some of higher MW sugars. I would also half your injection volume to 10ul if you can go lower and still get good detector results you will see better separation as well.
One other thing I learned the hard way which was mentioned earlier is the longer equilibration time HILIC needs which seems to increase the older the column gets or the rougher you treat it.
Some people may suggest EtOH or IPA which allows for a lower water % without sugar precipitation, however for long term analysis it seems to degrade columns and I get lower recovery % then I do with a 50/50 ACN/DI mixture. Still may be worth a try with your sample though never did seaweed before. Let us know how it goes.

Good luck