Related Substance of Etofenamate

[jayroseville58]

Instrument HPLC
Column ODS HYPERSIL (mm) 100x4.6, 3µm
Buffer pH 8.0 3.25g Tetra-n-butylammonium hydroxide, 40%w/w +
1.3g di ammonium hydrogen phosphate + 900mL water, pH adjusted into 8.0 with dilute phosphoric acid (afterwards dilute to volume into 1000mL with water)
Mobile phase Mobile Phase A: Buffer pH 8.0
Mobile Phase B: 100% Methanol (HPLC Grade)
Gradient Program
Time (minutes) %A(v/v) %B(v/v)
2 45 55
3 40 60
13 35 65
20 12 88
26 12 88
27 0 100
31 0 100
33 45 55

Flowrate 1.6 mL/min
Injection Volume 20 µL
UV Detection 286 nm
Column Temperature 40°C



I used the above mentioned hplc condition for the analysis of etofenamate's relate substance. At first, I was getting good result but as time goes by, my active's retention time changes dramatically from 11-12 minutes to 15-17 minutes. The sample formulation we had was gel type. Following the procedure below for sample preparation, we experienced peak shouldering. Apart from that, we experienced also prime overpressure, compromising the whole trial run. What could be the reason from the problems we faced lately. thanks.

SAMPLE PREPARATION: 1000mg of sample into 50mL vol. flask, diluted to volume with 70% methanol then sonicated for 30 minutes to dissolve. Sample was filtered with 0.45µm nylon filter prior to vialing and injection.

[tom jupille]

as time goes by

How much time? How many injections? And is the change gradual or abrupt?

[jayroseville58]

run time per injection used was 35 minutes. I have 9 standard injections plus four sample injections which gave good peak shape results. So, all in all after 8 hours. my injection run suddenly became not good (peak shouldering I mentioned earlier).

[jayroseville58]

as time goes by

How much time? How many injections? And is the change gradual or abrupt?

run time per injection used was 35 minutes. I have 9 standard injections plus four sample injections which gave good peak shape results. So, all in all after 8 hours. my injection run suddenly became not good (peak shouldering I mentioned earlier).

[upamaniah]

hi jayroseville,
your column pH range is 2.0 - 8.0 at 40 deg C at higher flow, you are running the system at higher risk for gel samples with minimum sample preparation, i request you to use Shodex Asahipak ODP50, you will get much better reproducible results.
for prime over pressure check the system without column using ACN:H20:: as mobile phase.

[HPLCaddict]

You're running a standard silica based column with phosphate buffer ph 8 at 40°C - that's at least at the edge of that columns capabilities. I don't know if your problems are related to this but you might experience pretty quick degradation of the silica...
Abrupt formation of peak shoulders often is the result of debris on the inlet frit. In that case, all peaks in the chromatogram should be affected. Did you try reversing the column?

[HPLCaddict]

Looking at your procedure again, I started to shiver .
Apart from being pretty high with pH, you're running a gradient with ion-pairing reagent, which is generally not such a good idea. But on top of that, you're going to 100% Methanol, which means you're flushing most of the ion-pairing reagent off the column. After that, your gradient doesn't even contain a reequillibration step! So there's just the time that your autosampler needs for the next injection (1-2 minutes?) for reequillibration, which is definitely not enough to reequillibrate with that mobile phase!
To be honest, I wonder how you were able to produce even a few runs with consistent retention times. Just by looking at that procedure I'm pretty sure that it CANNOT work.

Did you develop this yourself or did you get it anywhere else? Did it really work in the past?

[jayroseville58]

Looking at your procedure again, I started to shiver .
Apart from being pretty high with pH, you're running a gradient with ion-pairing reagent, which is generally not such a good idea. But on top of that, you're going to 100% Methanol, which means you're flushing most of the ion-pairing reagent off the column. After that, your gradient doesn't even contain a reequillibration step! So there's just the time that your autosampler needs for the next injection (1-2 minutes?) for reequillibration, which is definitely not enough to reequillibrate with that mobile phase!
To be honest, I wonder how you were able to produce even a few runs with consistent retention times. Just by looking at that procedure I'm pretty sure that it CANNOT work.

Did you develop this yourself or did you get it anywhere else? Did it really work in the past?

Actually, this method was just given to me for validation purpose. The method was adapted first on injection solution formulation. The purpose is whether it could be applicable on cream and gel formulation as well. My superior was trying to use longer length column (this time 150mm X 4.6mm) but we still experience increase on back pressure. Afterwards, we use two individual pump to connect into another HPLC system which can occupy even higher pressure max 6000 psi but still prime overpressure was present. My superior then decided to even use UPLC and pressure really reaches between 8000-11000 psi. My superior then go back again with the use of Alliance system (the one with internal pump for gradient use) but this time with even more long length column (from 150 to 250 mm length) and even large particle size from 3micrometer to 5micrometer. Problem then was on system suitability resolution.

On this point I think, what I really need is to know what development can I suggest for the method we were using. Or how can I able to convince my superior that this method really could not work base from the conditions we used.

Thanks really for the great help.

[tom jupille]

Actually, this method was just given to me for validation purpose. The method was adapted first on injection solution formulation. The purpose is whether it could be applicable on cream and gel formulation as well.

The short answer is "no". It's a horrible method to begin with, and the matrix is certainly not helping. This is one of those cases where you would be better off starting over. That said, the biggest issue would be sample cleanup.

[jayroseville58]

Actually, this method was just given to me for validation purpose. The method was adapted first on injection solution formulation. The purpose is whether it could be applicable on cream and gel formulation as well.

The short answer is "no". It's a horrible method to begin with, and the matrix is certainly not helping. This is one of those cases where you would be better off starting over. That said, the biggest issue would be sample cleanup.

so what could now be the best suggestion method i could use in order for me to be able to analyze the relate substance of my gel and cream samples? thanks

[tom jupille]

Same as with any method development project. First of all, contact the original developers to find out why they ran at a pH that is at the upper end of the stability range for that column and why they combined a gradient with an ion-pairing reagent (both are things to be avoided if possible).

Then get hold of a stressed API sample and run full-range gradient on a good, new-technology C18 column (Hypersil has been around for a loooong time) at a fairly low pH (around 2.5 is a good general choice) and a moderate pH (maybe around 7; both of those are attainable with phosphate buffers). Adjust selectivity as necessary (gradient steepness, temperature, organic solvent, pH, or a different column).

If you have a wide range of retention and you know or suspect that you have a mix of ionized and neutral compounds (comparing the results at two pH values will give you some insight there), you might consider a mixed-mode column instead of C18.

All of the above is the easy part . Now you have to work on getting your analytes out of the cream formulation. Extraction? LLE? SPE? are all possibilities. Start with a matrix blank and run your chromatography method to look for interferences (tweak the cleanup as necessary) then move on to a blank spiked with the stressed API and look at recovery (again, tweak the cleanup as necessary).

[upamaniah]

Dear Jayrose,
try to use XSelect CSH C18 with 100mm L x 4.6mm ID x 3.5 um (Waters New chemistry column)
also i suggest you to add, 5 - 10% ACN in mobile phase B, and restrict your gradient proportion to 90% of B
good luck
upamaniah

[jayroseville58]

thank you for the effort on suggesting plausible solution on my problem. regards all.

[Vlad Orlovsky]

When I read something like this I am just wondering if people even know why they are adding stuff to the mobile phase (buffer, acid, IP, etc.). There is a reason why each component is added to the mobile phase and why certain gradient is applied. What are mechanisms of retention, what is achieved by doing specific mobile phase composition. The molecule is basic in nature and mobile phase contains IP reagents for acids. This should be a straight method on any of the columns, even reverse phase will work wit the gradient at lower pH (since some of the impurities are acidic and zwitter-ionic in nature. Down the road people who will move this to validation will call column manufacturer complaining why columns are not working, why method is not reproducible and blame everybody for the shortcomings of the method.

[jayroseville58]

thanks for added concerns. meanwhile, we made a detailed report about the scenarios that we experienced during our trial runs. I also referred all the suggestions mentioned earlier to my supervisor but he was not open in applying all of those. Instead, he just want to return back the problem to the supplier of the method and let them do the job of giving us other ways of solving the problem. i'll just keep you all updated with the future experiments we might encounter. thanks.