SPE and HPLC of Amines

[Frosty]

Hello,
I am trying to replicate an assay for amines found in equine faeces from a paper by SR Bailey et al. (Research in Veterinary Science 74 (2003) 113-118) but am having problems getting the standards to run reproducibly.

The amines are first dissolved in water (or acetone) and derivatised with dansyl chloride. They are then extratced using C18 cartridges and eluted with acetonitrile and filtered. I am looking at concentrations in the range of 0.1 ug/ml up to 20 ug/ml. Heptylamine is used as an internal standard at a concentration of 5ug/ml.

The HPLC is fitted with a Discovery C18 column and a Discovery C18 supelguard column. Solvent A is 5% acetonitrile in water and B is acetonitrile. The gradient elution is 25% B at the start increasing to 45% B at 5 minutes, 60% B at 10 minutes and 70% B at 15 minutes. It is then held isocratically for another 15 minutes. A flow rate of 0.7ml/min is used and UV detection at 250nm.

The problem I have is that I don't seem to get the same peaks twice! If I run a sample containing a mix of 6 amines I get 6 peaks, but if I run the amines individually I don't get single peaks that correspond to those in the mixed sample! Also the peaks seem to move randomly. I noticed that some of the peaks seemed to be coming off very close to the end of the run so I then did longer runs of 60 minutes. I am also heating the column to 60degC (as this was suggested by another amine paper).

Can anyone spot what I am doing wrong? I'm a bit concerned that if I can't get the standards to run reproducibly there will be no point even trying to run samples from the horses! I'm not sure if the problem is with the SPE or the HPLC.

Any suggestions would be most appreciated!
Best Wishes,
Frosty

[SIELC_Tech]

As alternative you can use mixed mode chromatography. In this case you can analyze even small amines (ammonia, methylamine, etc.) using HPLC and ELSD/LC/MS detection. You don't need derivatization with this approach.

Please see links below:

http://hplcmethods.com/compound_005.php (ammonia and simple amines)

http://hplcmethods.com/compound_015.php (benzylamine and aminoacids)

http://hplcmethods.com/compound_007.php (underivatized amino acids)

You can look at other amines here:

http://hplcmethods.com/Applications_By_Compound.php

[Mark Tracy]

Non-reproducible retention times can be caused by
- pump problem such as worn out seal, leaky fitting, bad checkvalve, bad proportioning valve
- temperature variation
- mobile phase problem such as inadequate mixing, poor pH control, imprecise volume measurements

In your case, I think you have a pump problem.

The derivatization and SPE steps control the size of the peaks, and have almost no effect on retention.

[HW Mueller]

It could also be that your derivatization is faulty. Dansyl chloride with amino acids gave highly variable, etc., results here, we used FMOC and did much better. FMOC is also used with amines. If you can do this via fluorescence it is highly sensitive.

[Frosty]

Hello,
Thanks everyone for your suggestions. I will check that the pump is working OK but I'm beginning to suspect it may be the derivatisation that is the problem. The only samples that will run with any degree of reproducibility are three mixed standards, all the individual ones either have no peaks or really horrible blobs! This makes me think that maybe the derivatisation has only worked for these three samples. I have carried out the derivatisation and SPE a number of times on all the amine standards but I'm really not sure it's working.

HW Mueller, could you please post some details of what to do with the FMOC? I've had a look on the web but most info is on protein synthesis and stuff. Do you still have to use SPE after the derivatisation step?

Thanks,
Frosty

[HW Mueller]

What you saw was most likely derivatization of amino acids, you can use their protocols, or if you wish, I can dig out what we did next week.. You will have to do a cleanup before derivatization, of course, not afterward, except one should get rid of remaining FMOC with adamantyl amine or glycine depending on whether you want it to come late or near tm (to). One should be aware that there is always some ammonia around and water which get derivatized also.

[Uwe Neue]

Here is a completely different thought. If your samples are dissolved in acetonitrile, you will get peak distortions, if a large injection volume is used. If this is a possible source of the problem, inject less or dilute the samples with water.

[Cliff Mitchell]

Dansyl contains an amine on it as well, you probably need some sort of buffer to set the pH and ionization of your system. Your analytes are still acids (carboxylate from the amino acid) and amines (on the Dansyl group).

[Cliff Mitchell]

Oops, I see now that you are analyzing amines, not amino acids. Still, your analytes are still amines and a buffer in your mobile phase should help

[Cliff Mitchell]

Here is a link to a previous discussion where I posted some stuff on FMOC derivatization

http://www.sepsci.com/chromforum/viewto ... light=fmoc

[Frosty]

Hello,
Thanks everyone for your helpful suggestions. I now have a few things to look into and try out! The derivatisation I am using just now includes a borax buffer at pH 10.5. If I have buffered during the derivatisation do I still need to buffer the mobile phase? (Sorry if that's a silly question, I'm quite new to HPLC). Also what would you suggest for sample cleanup?
Thanks,
Frosty

[HW Mueller]

The reason for buffers was discussed extensively before. Quickly: If you don´t want to have the pH of the mobile phase to jump all over you use a buffer (to keep ionizable analytes or possibly also silanols in the same state). A buffered sample, can, therfore, if its pH is quite different than that of the mobile phase, be a greater problem than an unbuffered sample.
You mean cleanup after the derivatization? I answered that above for FMOC. Are you still using Dansyl?