How standard degradation can affect sample recoveries

[mohan_2008]

Hi guys,

This might be a stupid question. Owing to a lot of interpretations and answers - I would like some expert advice.

My friend Kaylash was running a standard curve (a three point) which will be used to quantitate the specific compound in the unknowns.

However, later to his dismay he found that the standards were degrading.
He saw a degradant peak just infront of the analyte peak, which consistently seem to increase with time.

Now my questions:

1. How do you confirm that the peak infront of the analyte is infact the degradant.
2. How is this standard degradation affect your sample quantitation. Is it going to increase/decrease the percent recovery of the analyte.

[Consumer Products Guy]

If your standard degrades, it will produce less area/signal for the concentration you "think" it is (the concentration you made). So in relation to that, the area/signal in the freshly made sample will be larger, relatively, and your results will be high.

If the "new" peak appears over time, and is not present in your new standard (same lot of solvent), then your original standard has degraded some, and that's your degradation peak.

[Bruce Hamilton]

If the standard is degrading, then the method sucks and should be fixed. How do you know the samples aren't also degrading?.

Just because a new peak appears, that doesn't immediately invalidate the calibration. The issue is whether the analyte peak area in the standard is also changing with the degradation.

It's easy enough to test for, just analyse an old and new standard of the same nominal concentration, and focus on the peak area difference..

As CPG notes, if the standard area drops with time, the sample results will be reported high, along with the assumption that the samples haven't also degraded.

Bruce Hamilton

[mohan_2008]

Thanks guys,

I agree with both of you.

This method was developed by my colleague. Due to a restriction with time, he has to develop this method haphazardly and forgot to point this standard degradation issue.

Samples were made in special formulations, which were found to increase the stability of the API in the formulation, as based on a 15 day stability study. Fresh standards were made on a daily basis to ascertain the stability timepoints.

Hence, the possibility of sample degradation can be ruled out. I was of the primary opinion that standard degradation can create a positive bias, or high error - however, wanted to confirm the same.

Thanks again for your prompt input.

[Bryan Evans]

Did he not check the stability of standard solutions during validation?
He should've checked those on the same run as sample stability
(just treat the 15 day old standards like samples in the sequence file,
bracketed by fresh standards).

If the standard degrades, you can try some of the following:

a). Use actinic glassware
b). Use actinic HPLC vials
c). Cover the autosampler with aluminum foil
d). Chilled autosampler

You have to be careful with option "d" to make sure nothing crashes out of
solution at lower temps.

[Jade.Barker]

d). Chilled autosampler...

We use a chilled autosampler and still have a change in the standard, it is sneaky because it takes a few days to take hold but then goes suddenly.

Our standard has 4 sugars, 2 acids and 2 alcohols:
Day 1, standard and check standard match,
Day 2 & 3 still Pass.
Day 4 (4th injection into vials) Ethanol drops outside the 5% range.
My guess is the multiple punctures in the cap eventually let the Ethanol evaporate since it is the only part that begins to fail.

[Bryan Evans]

So maybe just keep the standard solution in the fridge,
and pour the stability standard solution into fresh new vials / caps
for each run?

Or some variation of that (recapping, different septums, ect.)

[Jade.Barker]

So maybe just keep the standard solution in the fridge...

Sorry, I didn't mean to be confusing, it is fine as long as the cap has not been punctured. We are able to store the standard in vials in the fridge as you say. We just need to use a "fresh" vial every day.

I meant to say the chilled autosampler won't always protect the standard, in our case only one component changed, it would be easy for a tech to miss since the Chromatogram still looks "right".

[mohan_2008]

Thanks Bryan,

I checked with him again. He did check the "solution stability" for 8 days, which is part of the validation protocol.

However, this standard degradations started to appear in the very recent studies and did not occur during the validation part.

We are suspecting a thermal degradation (which is frequent with our compounds). Light sensitive materials typically form a pale yellow or pale coloured solutions typically.

We will keep on searching for the possible causes. Also, the diluent itself may degrade the material which then has to dealt with.

[danko]

Hi Mohan_2008,

Samples were made in special formulations, which were found to increase the stability of the API in the formulation, as based on a 15 day stability study. Fresh standards were made on a daily basis to ascertain the stability timepoints.

Why not prepare the standard solution/s just like the formulation guys prepare the product/samples (i.e. using the ingredient/s that increase the stability of API)?

Best Regards

[mohan_2008]

Thats a good idea Danko.

However, the formulations tend to be very viscous to shoot directly onto the hplc system, and certain formulations tend to stick to the column making it hard to rinse them out.

Hence, we resorted (as is the procedure) to heavy dilutions with a solvent such as water/ACN or water/MeOH or as appropriate.

[danko]

Hi Mohan_2008,

It sounds like yet another rationale for emulating the samples.
From my point of view it is quite simple: If you can handle the samples then you should be able to handle standards that resemble the samples too. Otherwise you can’t really trust your results.

Best Regards