Fluorescence response problem

[Malarac]

Hi all,

I have a fluorescence issue that has me stumped and I hope someone on the forum can enlighten me.
Firstly, the setup:
The system is Dionex 3000 system with a Waters 2475 FLD (inherited from another lab).
I am looking at amino acids (glutamate and GABA) using precolumn derivatization with OPA + mercaptoethanol (all done in the autosampler) and run an isocratic MP (25% ACN, 1% THF, and 74% 50mM Na acetate buffer at pH4).
Detection settings: Ex = 337 nm, Em = 454 nm.
I use an internal standard that is added to the sample (by the autosampler) prior to the derivatization procedure.

A single run takes about 25 min and I run about 40-60 samples at a time with standards at the start, end and in the middle for the longer runs. My problem is that during the run something changes in the response of my samples (standards and unknowns). Basically, my response can be one of two types. 1) a low response from all analytes and 2) a much larger response so that my signals are close to maxing out (at a gain of 1). This wouldn't necessarily be a problem but the change in response is analyte dependent. For example, compared to the internal standard response the relative response of other analytes can be the same, lower or higher. Obviously this messes up any quantification.
I realize that your first thought is reproducibility, however, I have consistent responses for 25 samples in a row (response 1) and then all of sudden the response changes and the next 20-30 samples are consistent (response 2). It is like a switch has been flicked.
My thought is that it must be the detector because a change in wavelengths could alter the total response and the sensitivity to each analyte. Could it be that the detector is flicking between two different Ex/Em settings? how could this happen?

I have an example picture of two comparison standards but I am not sure how to attach it. I will happily email it to anyone if it would help or attach it if someone fills me in on how...

I hope I have explained this sufficiently.
Thanks,
James

[danko]

I have consistent responses for 25 samples in a row (response 1) and then all of sudden the response changes and the next 20-30 samples are consistent (response 2)

Hi James,
Is it sample dependant like this: If you run the same 25 (the fist 25) samples once again you'll get the same response as you got during the first run?
Or will the response change after the first run regardless the first results/values?
What will happen if you reverse the order of injection - i.e. if you inject the last samples first?

Best Regards

[Malarac]

The effect happens with my standards so it is not sample dependent. For example, I make up my standards then add 10ul of each to my vials. The same standard solution (although in 2 different vials due to the derivatization procedure) run at the start will give a different response to the one at the end of the run... I don't believe it is the autosampler given the variable differences between analytes so there is no way altering the amount of mixing/addition could affect one analyte differently to another in a combined standard.
I have even rerun samples the following day to try and get all of my samples to respond the same, so I have examples of some samples with both responses. However, it is not always response 1 first then response 2, the order can be switched (response 2 first then switches to response 1) so it does not appear to be an 'over time' effect either.
James

[danko]

If you only transfer 10uL in each vial (I assume you use some kind of small volume vials) I can't imagin you can re-inject from the same vials. How much do you inject by the way?

If I were you I would have tried injecting from the very same vial several times to see if something happens over time.
I suspect nothing will happen - you'll end up with nice similar responses.
My guess is the derivatisation you're having trouble with.

Best Regards

[Malarac]

Hi Dancho,

I actually inject 1 ul after in vial derivatization. Due to loss of fluorescence over time I can't re-inject from the same vial (I lose signal). My sample prep is as follows:

Begins with 10 ul sample (acidic). The autosampler then adds 10ul of internal standard (acidic) and mixes, followed by 20ul of borate buffer (basic pH 10) and mixes, followed by 5ul of OPA reagent (OPA and mercaptoethanol in 90% methanol, 10% borate buffer) and mixes. The sampler injects 1ul after a 45 second wait.

I would have thought that if the derivatization was the problem it would vary from sample to sample rather than in blocks of samples at a time?

Thanks,
James

[danko]

I thought if you re-inject - despite the fluorescence loss (decrease?) - you migth experiance the same ratios between the (response1) and (response 2). If so, then it would be something with the derivatisation to do(?).
How is the mixing of the different components performed by the way ?

Best regsads

[Malarac]

That could work but the only way to tell was if I happened to get lucky and catch the system when it switches. I could try this with all samples and hope...

Mixing is performed by drawing and dispensing within the autosampler needle.

James