Chromatography Forum

Hi! my name is Pamela Arias. I am trying to validate a method to analyze aminoacids in foods by HPLC.
Right now I am injecting standard solutions of diferents aminoacids, and I dont know why the peaks dont appear. I have tried changing diferents variables but it doesnt work.
I am using 2 buffer solutions:
A: sodium acetate trihydrate + triethylamine + EDTA + pH=7.2 + tetrahydrofuran
B: sodium acetate trihydrate diluted + pH= 7.2 + methanol + acetonitrile
I am using a stock solution of OPA (50 uL) + Borate (250 uL) + sample (50 uL) + FMOC (50 uL) and I am injecting 100 uL.
Everything is microfiltered before to inject.
The column C18 I am employing is 25 cms.
I am using a flow gradient for the buffer solutions like this:
Time (min)= 0,17,19.5,20,24,24.5,25
%B= 0,60,100,100,100,100,0
Flow(mL/min)= 0.8,0.8,0.8,1.3,1.3,0.8,0.8
The detectors I am using are UV (338nm) and FLD (340 - 450nm)
I dont know why I am doing wrong, this is my thesis of pregrade. I am in the last semester of chemistry in the University of Antioquia in Medellin-Colombia. If you know something that I can do to improvement my method, I would be very grateful.
Thanks.

http://www2.shimadzu.com/apps/appnotes/ ... oacids.pdf

UV detection is poor due to lack of chromophores and fluorescence detection requires pre or post colum derivatisation.

It appears that you are using an OPA/FMOC precolumn derivatization procedure. Are you following an established technique from the literature? Sample preparation and derivatization of amino acids can be tricky. Also, does your HPLC system perform properly if you run a test compound?

UV detection is poor due to lack of chromophores and fluorescence detection requires pre or post colum derivatisation.

That's what the OPA is all about!

Pamela, the first question is "are you running an existing, validated method, or are you trying to develop your own procedure?"

If you are running an existing method, then it *should* work. Possible reasons for failure:
- the derivatization reaction with OPA is failing
- the derivatization is working, but the derivatized AA's are decomposing before chromatography
- the derivatized amino acids are not eluting from the column
- One or both of your detectors are malfunctioning.

If you are developing your own method, many of the same failures are possible, but troubleshooting will be more difficult, because you do not have a "known" response to compare.

If this were my problem, I would address the issues one at a time. For starters, work with only one amino acid!

1. Derivatization reaction. Remove the column and connect your injector to your detectors with a piece of capillary tubing (1 meter of 0.25mm id tubing should work) in place of the column. Prepare your sample per usual and inject. You should see a rapid "spike" response from the both detectors. If you do *not* see a response from the fluorescence detector, that means that either the detector is malfunctioning or that the derivatization reaction has failed. In that case, run the detector diagnostic tests as recommended by the manufacturer. If the detector checks out OK, then the problem is your derivatization reaction. I'm afraid I can't help on that, beyond suggesting the obvious: pH, temperature, reagent purity, etc.

2. If you *do* see a response with fluorescence, that tells you that the detector and derivatization reaction are working. Now prepare a fresh sample and repeat the injection every 10 minutes or so for several hours, while tracking the response. That information will tell you something about the stability of your derivative in solution. If it is decomposing quickly, then you will have to make sure that you prep a sample immediately before injection.

3. If sample stability is not an issue, you can re-install the column. Make an injection and run a simple full-range gradient:
A = your aqueous buffer + 5% acetonitrile
B = 100% acetonitrile
Flow = 1 mL/min
0 - 100% B in 20 minutes.

You should see a peak come out *somewhere*. It may be ugly, but it should be a peak. If you do *not* see any peaks, that suggests that either your buffer has been prepared incorrectly or your column is very, very dead.

4. If you *do* see a peak, that means that your derivatization, detectors, and column are OK. That suggests that your gradient conditions are bad. You did not specify what the % organic solvent was in your A & B solutions. For reversed-phase gradient HPLC to work, the "solvent strength" (approximately equal to the % organic solvent) should increase during the run. Identify what the % organic is when your peak elutes from the column. Now go back and look at your original method to check that you are going to at least that high a % organic. If not, that is your problem, and you will have to modify your method accordingly.

5. Once you have gone this far, you can add additional amino acids and figure out the best conditions for your separations.

This derivatization mixture looks weird. Wherefrom did you get this procedure?

And what was the gradient developed for? Are you using the same column as in the original procedure?

The OPA/FMOC methods that I have seen do not use THF in mobile phase A. What is the % of THF that you are using? Your mobile phase A elution strength might be too high...

a) the use of OPA and FMOC at the same time looks also weird to me.
I would use OPA or FMOC, otherwise I would expect to receive a mixture of both derivatives from each amino acid.

b) I miss the thiol (mercaptoethanol) compound for the OPA derivatization, otherwise the reaction won't work

c) is the pH of the borate buffer correct for the derivatization reaction(s)?

Also, FMOC derivatized amino acids are available in high purity, to use them makes it easy to get the HPLC to work and to check on the efficiency of the derivatization.
I said something like that before: I admire anybody who can get the OPA method to work, we got a mess and gave up in disgust.

For those that find the bundle OPA/FMOC weird, it exists and used due to the inability of the OPA to derivatize secondary amino groups (i.e. proline etc). They start by reacting with OPA, a very fast reaction, and then use FMOC to derivatize mainly proline and it's derivatives. They are able to resolve all derivatives so it is all good.

Having said that, I have heard a lot of people having problems with OPA... Most success stories comes from groups that are using automated derivatization by using the autosampler...

You can have a look at this method. Some of the universities here in the U.S. have started to use it:

Biomedical Chromatography. Volume 18. Issue 9. Pg.s 630-636. (Nov 2004)

A fully automated amino acid analyzer using NBD-F as a fluorescent derivatization reagent
Chiaki Aoyama 1, Tomofumi Santa 1 *, Makoto Tsunoda 1, Takeshi Fukushima 1 2, Chieko Kitada 3, Kazuhiro Imai 1 4

A fully automated amino acid analyzer using NBD-F (4-fluoro-7-nitro-2,1,3-benzoxadiazole) as a fluorescent derivatization reagent was developed. The whole analytical process was fully automated from derivatization, injection to HPLC separation and quantitation. The derivatization reaction conditions were re-evaluated and optimized. Amino acids were derivatized by NBD-F for 40 min at room temperature in the borate buffer (pH 9.5). The derivatives were separated within 100 min and fluorometrically detected at 540 nm with excitation at 470 nm. The detection limits for amino acids were in the range of 2.8-20 fmol. The calibration curves were linear over the range of 20 fmol to 20 pmol on column with the correlation coefficients of 0.999. The coefficients of variation were less than 5% at 3 pmol injection for all amino acids. Amino acids in rat plasma were determined by the proposed HPLC method.

FMOC-Cl derivatizes both secondary and primary -NH2. I got disgusted with the OPA method as I ran out of patience trying to lower the side reactions. There are at least 3 molecules that have to come together, usually the yields of such syntheses are not so hot.
To a certain extent it seems a matter of attitude and also of what one wants to do.