Chromatography Forum

I have difficulties in separating my sample.
My sample is an aqueous extract of plant and it is very water soluble.
The mobile phase i'm using is NH4H2PO4:ACN = 70:30, pH 5 with flow rate of 0.6 mL/min. The column that being tried is silica column.
However, the separation of peaks is not well.
Any suggestions to make it better ?

I have difficulties in separating my sample.
My sample is an aqueous extract of plant and it is very water soluble.
The mobile phase i'm using is NH4H2PO4:ACN = 70:30, pH 5 with flow rate of 0.6 mL/min. The column that being tried is silica column.
However, the separation of peaks is not well.
Any suggestions to make it better ?

please describe your sample in detail ,structure ，properties……I think more people can help you !

Ok, thanks for your opinion.
My sample is an aqueous extract of plant.
My job is only to get the peaks separate nicely, and so far there is no reported journals about the detail contents of the plants, unless tannic acid, but this is a very common component in a plant.
Therefore its quite headache for me. Hope that someone can really help me..thanks a lot!!!

This sounds like a textbook HILIC problem.
For an introduction to HILIC please se our guide to HILIC at
http://viewer.zmags.com/publication/95d ... 95db93bd/1
or visit our website at
www.sequant.com

If want to use the Hilic column, what is the preferred mobile phase for my sample?

Further on the HILIC approach:
The following paper describes analysis of plant extract metabolites in great detail, including chromatograms showing the elution position of different classes of compounds: V.V. Tolstikov, O. Fiehn, and N. Tanaka, Meth. Molec. Biol. 358 (2007) 139-156 (Humana Press). They used a HILIC capillary (PolyHYDROXYETHYL A) for the polar metabolites and a C-18 monolith for the nonpolar ones. The gradient involved a decreasing ACN gradient, starting at 85%. Since they were using ESI-MS detection, they had ammonium acetate in the mobile phase.
In your case, I presume you want to use absorbance detection, since you're using a phosphate salt (which is transparent at ~ 215 nm). Instead of ammonium phosphate, use triethylamine phosphate (TEAP); it's appreciably more soluble in HILIC solvent. I recommend starting with 85% ACN containing 15 mM TEAP overall. Make a stock solution of TEAP by dissolving a weighed quantity of H3PO4 in water and adding triethylamine until you get the pH you want, then filtering and diluting to the mark in a vol. flask.

thanks all, but the column that i currently using is silica column....

Any column material that's at least somewhat hydrophilic can be used in the HILIC mode as long as you have enough organic solvent in the mobile phase. Silica columns are adequate for undemanding applications such as analysis of drugs. For a really complex mixture like a plant extract or a protein digest, you should use a material with a reasonably thick, well-hydrated coating. Examples include Merck/Sequant's ZIC-HILIC, Tosoh's TSK Amide-80, and PolyLC's PolyHYDROXYETHYL A. I don't think you're going to succeed in your analysis unless you buy a more suitable column.

With silica column you are operating in HILIC mode. You are perfectly fine with the column you have and don't need to buy another column if you don't have a budget for it. In your case desired selectivity can be achieved by playing with amount of ACN, buffer nature, buffer concentration and buffer pH. You need to start with higher organic, in order to achieve retention in HILIC mode you need at least 65-70%. In your case it might be complicated , because you probably have water extract. I would try to do wide gradient from 80% of ACN to 40% ACN, and buffer concentration from 5 mmol to 20 mmol over 20 minutes. You can also try to play with pH of the mobile phase. pH will change ionization state of your compounds (if the are ionizable within this pH range). It will also change ionization of silanols, causing change in cation-exchange properties of silanols. You can also try to switch to a different buffer (ammonium formate or acetate), but remember that UV cut-off will go from 200 for phosphate buffer to 220-230 nm for formate buffers. Also you need to look at solubility of your compounds at high ACN concentration and watch for buffer concentration at higher organic (you don't want buffer to crush out of MP!!!).

You can call me on Monday to discuss your problem (see phone number at our website). If you can send me a sample we can do fast screening in our lab free of charge

any chance of showing us a chromatogram of your application?
we have no idea what is really wrong otherwise
how many peaks
retention times, capacity, resolution

the simplest problem with HILIC applications like yours that give really bad chromatography for example is that the sample is made of too much water and this disrupts the separation.
another thing is the buffer concentration, the more the less ionic interactions, but is it necessary? what is the exact buffer composition right now?

Silica column is unfortunately the worst type of column to do HILIC with, because the liquid-liquid interactions are done on a very small volume of the particles and it is less stable then other stationary phases for this. it is also a lot more susceptible to detonation by the water that can simply stay stuck in the pores of the silica.
and in HILIC mode you are forcing the water to be next to the silica

anyway a picture of what is going on will save many needless writing on our part and give a far better understanding of what is going on

Injecting a 100% aqueous extract will cause problems in HILIC. Dilute your sample 1:2 with ACN prior to injection (i.e. 66% ACN in injected sample solution) and use 80/20 ACN/buffer (isocratic) as your mobile phase. Silica columns are fine for HILIC. If you have too much retention you can reduce the %ACN or use a gradient from 80% ACN to 50% ACN.

unmgvar, could you expound on this?:

"Silica column is unfortunately the worst type of column to do HILIC with, because the liquid-liquid interactions are done on a very small volume of the particles and it is less stable then other stationary phases for this. it is also a lot more susceptible to detonation by the water that can simply stay stuck in the pores of the silica.
and in HILIC mode you are forcing the water to be next to the silica"

Why would liquid-liquid interactions be only on a very small volume of the particle? What does less stable mean here? What does "susceptible to detonation" mean? I did experiments with T2O and saw no evidence for water to be stuck in the pores, what is your evidence for that? How does the HILIC mode force water to be next to the silica?

I have difficulties in separating my sample.
My sample is an aqueous extract of plant and it is very water soluble.
The mobile phase i'm using is NH4H2PO4:ACN = 70:30, pH 5 with flow rate of 0.6 mL/min. The column that being tried is silica column.
However, the separation of peaks is not well.
Any suggestions to make it better ?

From the above text it would appear you are using a silica column in a reverse phase application. Silica columns are designed for normal phase operations with a low or negligable water content.

If I'm incorrect and what you are using is a bonded phase silica such as C18 or C8, ignore this.

Also for complex extract separations, you would be better off exploring a gradient rather than an isocratic LC method to improve separation.

I have difficulties in separating my sample.
My sample is an aqueous extract of plant and it is very water soluble.
The mobile phase i'm using is NH4H2PO4:ACN = 70:30, pH 5 with flow rate of 0.6 mL/min. The column that being tried is silica column.
However, the separation of peaks is not well.
Any suggestions to make it better ?

From the above text it would appear you are using a silica column in a reverse phase application. Silica columns are designed for normal phase operations with a low or negligable water content.

If I'm incorrect and what you are using is a bonded phase silica such as C18 or C8, ignore this.

Also for complex extract separations, you would be better off exploring a gradient rather than an isocratic LC method to improve separation.

Agreed we need more information though I doubt the OP is using a bare silica column. The OP is probably confused in that most LC columns are built off of Silica. What we need to know is what the bonded phase is.