Peak fronting in acetylsalicylic acid assay

[HW Mueller]

Xisiqomelir, waiting for the k.
Could it be that you missed what Uwe and Dancho tried to tell you, namely, that if you still use the same column and the flow rate is not even faster than before you "changed the procedure", than you may no longer be doing chromatography?

[Bruce Hamilton]

If I was the original poster, I'd be a little peeved at the implicit assumption that I don't know much chromatography.

I assume, given that he has an expensive UPLC-type toy, and he's developing methods for regulatory approval, he would move to a different column, and possibly even a different methodology.

My answer ( the original poster may differ ) to some of the above comments. Veterinary drugs are usually controlled by the same regulatory body that controls pharmaceuticals. Consequently, developing a method based on an established compendial procedures will usually attract less scrutiny. It may be that the purchased bulk API is tested using the USP method, rather than the EP method.

In the bad old days, when veterinary grade APIs were only about 1/3 price of pharma grade, one employer used to buy both grades, and back blend to 98.5% ( spec was 98 - 101.5 % ) assay or upper limit of an impurity, then full retest of the blend to compendum before making tablets.

The regulators these days wouldn't allow that. Besides, veterinary grades can be as expensive as pharma grades.

With regard to the continued use of an ion-pair agent ( thanks CPG ), the method works, and presumably was originally developed and approved by analysts who found it met all objectives. Modified, it may even meet the OP's needs as well, despite the scorn of some...

Yes, knowledge and columns have improved, possibly offering superior options, but that does not mean the original chromatographers were a bunch of dumb bunnies randomly trying to achieve separations.

Please keep having fun,

Bruce Hamilton

[HW Mueller]

Bruce, everybody here knows what they are doing???

[Bruce Hamilton]

Bruce, everybody here knows what they are doing???

Unless the original poster is God ( who may have been experiencing an off-day in heaven's laboratory ), he/she doesn't equate to everybody here.

Please keep having fun,

Bruce Hamilton

[cyanogen78]

Yes, it seems that the peak is very close to t0...so it is not the solution of the problem.
I think fronting emerges due to "wrong" mass transfer - probably due to competition between molecules of the IP reagent and the formed ion pairs themself. So if this is the case - the solution could be tedious to reach if following the same direction. The concentration of IP reagent could be lowered and peak symmetry checked. The temperature could be also a factor. If all instrumental conditions are checked and there is no improvement - probably the reason for peak fronting in this case has thermodynamic nature, and is in mass transfer between the analyte, mobile phase, IP reagent and stationary phase.
I think the easiest solution is to forget about ion pairing and to work under aspirin pKa with highly aqueous mobile phase. Of course suitable column is essential.

[HW Mueller]

What ion pairs??

[Bryan Evans]

Below is separation of acetylsalicylic acid and salicylic acid on Cadenza CD-C18:
http://www.imtaktusa.com/site_media/fil ... TI173E.pdf

[ccuillie]

http://www.imtaktusa.com/site_media/fil ... TI173E.pdf

This spectra, all be it very nice looking, is scanning the wrong wavelength for acetylsalicyclic acid and salicyclic acid.

acetysalicyclic acid has a max absorbance at 278nm and salicyclic acid has a max abs at 302nm.

[Consumer Products Guy]

http://www.imtaktusa.com/site_media/files/technical_information/TI173E.pdf

This spectra, all be it very nice looking, is scanning the wrong wavelength for acetylsalicyclic acid and salicyclic acid.

acetysalicyclic acid has a max absorbance at 278nm and salicyclic acid has a max abs at 302nm.

1. There's no "wrong wavelength" if one validates the test method using those conditions. I think the supplier just chose a compromise wavelength to demonstrate the separation. In real life, one could use detector wvelength switching to be at wavelength maximum for each, but the assay is simpler in this case (large amount of analyte, few interferences) if a single compromise wavelength is used. We have several assays here that do similar.

2. I'm not sure the original poster needs to have or use a validated test procedure for pet pharmaceuticals, I have no idea. Are there regualtions for these as well? A Pet Pharmacopeia????

[Bruce Hamilton]

The USP offers veterinary drug monograps, which usually refer back to the human drug monographs in the USP/NF. The curious can download them without a USP subscription but, surprise, surprise, you would have to pay for the USP to obtain the detailed monograph.

The British Pharmacopoeia 5 volume pack includes a separate volume called British Pharmacopoeia ( Veterinary ) 2009.

As I noted above, it's fairy usual for the human drug regulator to also regulate veterinary drugs. These days, the quality and raw material prices are similar, if not identical.

That's probably why the oroginal poster wanted to start with ahuman drug method..

Please keep having fun,

Bruce Hamilton

[mijnadarian]

Hello all

Had a problem with Salicylic Acid assay and thought maybe I can get your feedback.

We have a product, that has 2.8% Salicylic Acid in it- The method used to run on a 30cm Supelcosil Supelco RP C18 x 4mm x 5 um column.
My predecessor said that, the method was running on 1 ml/min @ 280 nm @25C
- the mobile phase was 90% MeOH and 10% (0.1%) TFA in water in Isocratic mode.

For some reason the method became useless after the assay was giving us high numbers.
Peaks became broadened, and the system suitability testing was failing-
Henceforth; I tried to develop a new method.

Recently , I have been trying to develop a method on the new Hilic Kinetex column by Phenomenex-
I'm running the assay on the 100mm x 2.6 micron x 4.6 mm
My system is a Tower P680 Pump by Dionex

I have tried all yesterday to get a decent peak shape, from the whole morning runs- They came no good- I was told that I can run this, with (Ammonium Formate 20mM) and ACN in a gradient method- 80% Buffer and 20% ACN @ 210 nm.
( I will explain the gradient programming later)


The Salicylic Acid peak was coming out at 1.65 minutes, with very low theoretical plate counts with rather broad peaks. The assay was ok in ballpark of 2.8%, which it should be. The R2 value was 0.998 with near perfect calibration.

..Still not satisfied with SST results, I tried to read some of the postings by you guys-

So, I changed the program and made a buffer of Potassium Phosphate 50mM and acidified it with 1ml of TFA- It brought down the pH to 2.6 range.. And for the Organic Phase, I used Acetonitrile and I am doing this in Gradient mode with following program;

At 0 minutes 20% ACN and 80% 50mM of Phosphoric Acid with 1 ml of TFA in 1 Liter
At 5 minutes yielding to 50% ACN and 50% buffer
And at 10 minutes- getting back to 20% ACN and 80% of buffer

The result was the following:
My Salicylic Acid peak became significantly in better shape with much much better theoretical counts-SST passed in all categories and the retention shifted from 1.6 minutes to 4 minutes-

But here are the two problems

1-My base line has a negative drift right after the end peak time and starts a positive direction drift at 7.5 minutes - I'm guessing this is normal with gradient elution?
Am I correct? FDA will be ok with the baseline negative- positive drift? It is not a huge drift- But I never had a drifting problem before-

2- I got low assay results- in 2.4% range- I have a feeling, if I run this on Monday-
I will get lower numbers again-

I must address; that my area counts and peak heights were different in the last analysis , but I used samples made from the morning. Does this matter?

Anyway,
I'm just thinking is there any better method to run the Salicylic Acid Assay or maybe I have to play with the mobile phase with the current conditions that I have worked out.
I'll be all eyes and ears, and I would like to thank all of you in advance for any word of wisdom..

Thank you all
Sincerely-
M.

[grzesiek]

"For some reason the method became useless after the assay was giving us high numbers.
Peaks became broadened, and the system suitability testing was failing-
Henceforth; I tried to develop a new method. " - why did you try to develop new method and not troubleshoot the old one?

"buffer of Potassium Phosphate 50mM and acidified it with 1ml of TFA" - why not phosphoric acid?

"was coming out at 1.65 minutes" - k?

"with very low theoretical plate counts" - how low is low for you?

"I'm guessing this is normal with gradient elution? " - post the picture

"FDA will be ok with the baseline negative- positive drift?" - they should not be interested at all

"but I used samples made from the morning. Does this matter?" - different samples give different results, not strange at all, why not use a standard solution?

I don't really get the idea what you did first what next etc. could you write the changes chronologicaly?

[mijnadarian]

-Thanks for the reply
why did you try to develop new method and not troubleshoot the old one?

First Attempt-
When I encounter the problem; I did try to change the mobile phase to 65% MeOH and 35% (0.1%TFA) and I got results ~ 2.8% which where I was looking for- Except the peek was tailing-
So, I thought since - there was no documentation about the column usage ; I assumed perhaps column is not reliable and it is aging and started wondering about; if the data was even reliable-
how low is low for you?

Second Attempt-
I was getting plates in low 200s- This is what I had; when I attempted with the new column from Phenomenex.( see the picture below)
at 0 minute @ 20% ACN and 80% 20mM Ammonium formate
at 5 minutes yielding to 50% ACN and 50% Ammonium formate
at 10 minutes going back to 20% ACN and 80% Ammonium formate
I was getting 2.8% results here as well- Except; the plate count was low- Everything else in SST was ok-

The third attempt was with
The new mobile phase was 50mM Potassium Phosphate- acidified to pH of 2.6 and starting at 80% and ACN starting at 20%
Just like the one I explained above-Following the exact gradient pattern mentioned - the SA peak shifted to 4 minutes with excellent theoretical plate counts- No tailing either-
I don't have the new picture, but theoretical plate counts is in 7000 range.
Like I said; after the peak ends; there is a negative drift and then positive direction drift-
My only question at this point is; why am I getting low results?
(I'll try to post pictures later)
Thanks again

[grzesiek]

"I was getting 2.8% results here as well- Except; the plate count was low- Everything else in SST was ok-" - what is ok, and what is not? if you can get your LOQ where you want it, I wouldn't care much about the plates

[mijnadarian]

Hello
Just to give you an update;
I ran the analysis with the Phosphoric Acid- and; I got a good baseline- So, I think TFA has some UV absorbance. Good baseline, good chromatography and Theoretical Plates are in 10000 range.
Results are in 2.9% range instead of 2.8%. Is the higher result because of higher efficiency of the column?