proper use of amino column

[benzech]

I have posted before of my trouble getting a stable RT with the silica-based amino column I have used and searching the resources has led me to some basic questions:

1. how common is it to use the amino columns with reverse phase solvents (acetonitrile/water) vs. normal phase solvents (EtOAc/hexane)?What are the advantages/disadvantages to each?

2. Should gradients be used or just isocratic run?

3. If used in an ion-exchange capacity, does the ion-ion interaction dwarf the H-bond (normal phase) interaction?

I am using this column to isolate what I think is a very polar low MW carboxylic acid (structure unknown). currently I'm using a gradient method that begins the run in 98:2 Acetonitrile/water and slowly introduces 20mM ammonium formate pH 2.8 in water to the flow. At about 20% B the compound elutes. My thinkning is that high organic content at the start of the run will allow retention of the compound and increasing aqueous content will elute it (if normal phase interaction is the mechanism of retention) and if it is an carboxylic acid then lowering the pH by introducing the pH 2.8 buffer will protonate the acid and cancel out any R-NH3+ -OOC-R ion-ion interactions. Compound retains well and separation from the matrix (soil extract) is good but the retention time is always shifting around sometimes by several minutes. Let me know if you see any obvious flaws here y'all.

[Uwe Neue]

There is a basic problem with amino columns when used in an aqueous environment, and you may have run into it. If you do not use special precautions to bring them to a low pH, they will destroy themselves due to the fact that your pH in the column is high, around 9 or 10. Your shifting retention times may be caused by this.

Another possibility is a very slow equilibration of the column with your very low concentration of the ammonium formate/formic acid.

Put a piece of pH paper at the column outlet when you are in your pH 2 buffer and see what the pH paper says.

[Ken Tseng]

Most applications with amino columns are done with acetonitrile/water in normal phase (or HILIC mode as some would like to say).

Usually, you can start with isocratic runs to get a general idea of working experimental conditions. Then optimize them with gradient runs. It depends on your analytes and experimental goals.

The state of the equilibrium (free base ratio) depends on pH and ionic composition of the eluent. To see the effect of non-protonated vs. protonated amino groups on elution time and plate numbers, visit http://www.shodex.com/english/dc030234.html

In general, you almost always would want to use polymer-based amino columns. A typical silica-based amino column’s performance starts to degrade almost right away. http://www.shodex.com/english/dc030214.html

Shodex has both silica-based and polymer-based amino columns. But, you can avoid a lot of frustrations and wasted resources by start using polymer-based amino columns right away. The Shodex NH2P has pH range is 2-13, and compatible with 100% water, acetonitrile, and ethanol. http://www.shodex.com/english/cn04.html

Other manufactures that also provide polymer-based amino columns are Supelco, Grace, Phenomenex, Waters, and others

[benzech]

pH at the start of the run was 5. pH at 98% buffer was 3-4.

I see that silica based columns degrade rather quickly over time but can that explain the shifting retention times from day to day?

[Uwe Neue]

I appears that you are now in a state where the column is acidified. The changes in retention should become less. However, your final pH is not yet what you are putting in at the column top. My opinion: either you flush the column with a high buffer concentration, or you will need to wait longer for stable retention times.

[Ken Tseng]

If your peak shifting is one-way over time and in the same experimental condition, it is probably caused by column degradation.

[benzech]

Yes, Uwe I had also considered a flush at high ionic strength but have not yet run the experiment. I assume the thinking there is that high ion concentration would flush away any tightly retained anionic components of the matrix that adhere to the column surface, what do you think?

Ken, I have seen shifts in both directions of up to three and four minutes. I have seen a general downward trend in RT though that is consistent with your statement. By the way, in the chromatogram you linked to entitled Effect of Free BaseRatio of Amino Group on Plate Number something did not make sense to me, the table header says at pH 2.84, the freebase ratio is 32%, I assume it means 32% R-NH2 to 68% R-NH3+. If the pH of this buffer is around 3 and the pKa of the NH2 moitey is around 9 then the HH eqn. gives about a 1:10^6 ratio, right? What am I missing?

[Bryan Evans]

Hi Benzech -

I think your problems stem from one of 3 things:
a. Poor column lifetime
b. Insufficient sample cleanup
c. Insufficient wash step in gradient

It's true that silica-based NH2 phases die quickly when exposed to water.

That's what makes Unison UK-Amino so special:
"unequalled durability against water elution."
http://www.imtakt.com/TecInfo/TI304E.pdf

So you get the durability of polymer based column but the efficiency
of 3um silica.

To answer your questions of how NH2 phase works (using UK-Amino as an example):
a. normal phase (for non-polar compounds, hexane eluent, ect.)
b. normal phase (water is the strong solvent)
c. normal phase + anion exchange
d. (weak) anion exchange

Below is data that shows how retention works for UK-Amino
http://www.imtakt.com/TecInfo/TI316E.pdf

For your application, consider using AcOH or ammonium acetate / acetic acid
buffer. Whats nice about this is you can increase the
concentration of competing anion (acetate) without lowering the pH
too much.

[Ken Tseng]

[quote="Ken, I have seen shifts in both directions of up to three and four minutes. I have seen a general downward trend in RT though that is consistent with your statement. By the way, in the chromatogram you linked to entitled Effect of Free BaseRatio of Amino Group on Plate Number something did not make sense to me, the table header says at pH 2.84, the freebase ratio is 32%, I assume it means 32% R-NH2 to 68% R-NH3+. If the pH of this buffer is around 3 and the pKa of the NH2 moitey is around 9 then the HH eqn. gives about a 1:10^6 ratio, right? What am I missing?[/quote]

Shodex NH2P is a polyamine column.

[lmh]

Ken Tseng said: "Most applications with amino columns are done with acetonitrile/water in normal phase (or HILIC mode as some would like to say)."

Ken, would you mind elaborating on this? I thought that Normal phase described the situation where the analyte interacts directly with the column stationary phase, while Hilic described a situation where the column stationary phase interacts with water, retaining an aqueous layer, into which the analyte partitions.

Clearly these aren't quite the same mechanism. I'm not sure I entirely believe in pure Hilic as, if this mechanism were the whole story, all columns operating as Hilic columns would be identical (because they're all just water).

I also thought that water was death to normal phase chromatography, but I'm not expert on chromatography. I just get confused with the differences between amino columns used in different modes.

[Ken Tseng]

Thanks for asking this question. Let me do my best to answer.

HILIC stands from Hydrophilic Interaction Chromatography or Hydrophilic Interaction Liquid Chromatography. There is an on going debate on the HILIC mechanism, and what its acronym really stands for. But, let’s not get into this. From a practical point of view, HILIC is just a fancy name that describes a sub-section of normal phase when water is used in mobile phase.

Normal phase simply means that the stationary phase is MORE polar than the mobile phase. When chromatography as a technique was first discovered, it was done with polar stationary phase and relatively non-polar mobile phase. It was considered “normalâ€

[Noser222]

1. how common is it to use the amino columns with reverse phase solvents (acetonitrile/water) vs. normal phase solvents (EtOAc/hexane)?What are the advantages/disadvantages to each?

Just throwing this tip out for you...don't use EtOAc or you'll eventually make amides out of your amines.

[lmh]

Thanks Ken! So does that mean there is some hope for trying hilic/watery-normal phase on amino columns using a gradient of methanol versus aqueous buffer instead of acetonitrile versus aqueous buffer? I've already found that even the tiniest proportions of methanol in a mix completely destroys any separation on Phenomenex hilic columns (which are a poly-ol type column, I believe). In old Hilic terms, I thought I understood why (methanol can also hydrogen bond to the hydroxyl groups on the column and will therefore infiltrate the water layer, so there's nothing much for the analyte to partition between). Now I'm not sure I understand why methanol is a problem, unless it's competition for hydrogen bonding between the column and analyte versus column and methanol?

It'd be nice to open up the whole range of amino columns from all manufacturers, and be able to use a solvent I can buy without sacrificing all other consumables for a year.

[HW Mueller]

The problem appears to be elution strength. I have used EtOH extensively and even that is much stronger than ACN in a HILIC setting. Judging from that, my guess is that MeOH is close to water in elution strength.

On the term "HILIC": It was mentioned before that the "bible" on TLC, namely Stahl´s book, has examples of "HILIC" from the early sixties. Nobody seemed to think that this is anything other than normal.

[Uwe Neue]

You can use the alkohols in conjunction with acetonitrile. This will change the selectivity, which may be of interest. See the following publication:

E. S. Grumbach, D. M. Diehl, U. D. Neue, „The Application of Novel 1.7 µm Ethylene Bridged Hybrid Particles for Hydrophilic Interaction Chromatography“, J. Sep. Sci. 31 (2008), 1511-1518