Strange peak splitting

[Uwe Neue]

The common drugs that do this kind of stuff is the family of -prils, such as captopril, enalapril. Proline containing peptides do the same thing. I am surprised that you are not familar with it.

Examples of this are in my book on the pages 362-364. If you do not have my book yet, I think it is time to buy one....

[HW Mueller]

Uwe, that´s my physical organic upbringing, it tells me from the "back of my head" that a 5 membered ring amine should have an amine inversion barrier below 80kJ/mol, thus not have a measurable barrier to inversion at room temp. Just checked some stuff, then, and found that all the proline inversion inhibitions are seen only in the amidized amine function. This is in keeping with all my proline derivatives, which all have the "free" amine and no visible (in HPLC) isomerization.
Now, I just found that the thermodynamic isomer distribution of proline type amides (peptides) is near 20% for the cis and 80% for the trans. Looks like amaryl started out with mostly cis and got more trans at higher concentrations, because he heated more for the dissolution??
Oliver did some funny temperature variations on his samples after all?

[amaryl]

Uwe, that´s my physical organic upbringing, it tells me from the "back of my head" that a 5 membered ring amine should have an amine inversion barrier below 80kJ/mol, thus not have a measurable barrier to inversion at room temp. Just checked some stuff, then, and found that all the proline inversion inhibitions are seen only in the amidized amine function. This is in keeping with all my proline derivatives, which all have the "free" amine and no visible (in HPLC) isomerization.
Now, I just found that the thermodynamic isomer distribution of proline type amides (peptides) is near 20% for the cis and 80% for the trans. Looks like amaryl started out with mostly cis and got more trans at higher concentrations, because he heated more for the dissolution??
Oliver did some funny temperature variations on his samples after all?

These chromtograms are for standard glimepiride (without subjection to any stress conditions).

Thanks and Regards,

Amaryl.

I am she .

[HW Mueller]

Sorry about that, but "he" covers everything for some of us old guys....

If it is the isomerization mentioned by Uwe it´s "thin skinned" and doesn´t take stress, as you two apparently showed again.

[unmgvar]

amaryl,

i went over your numbers.
i took out chromatogram 2 from the top out because of the disturbance and my little number crunching makes me conclude that those two peaks are not isomers:

you changed concentration by a factor of 2 each time (so except for my step from chrom 1 to 3 which is 4 all other steps are also 2)

these are the conc steps:

4.005128205
2
2.000640205
2
2


these are the steps displayed by peak the peaks when i looked at the changes displayed in the are:

peak2 - peak 1 - combined areas
4.180894031 - 0.80515081 - 1.488550764
2.039990666 - 1.54845357 - 1.827944263
2.190746685 - 2.431857275 - 2.278856874
1.862108755 - 1.064632479 - 1.551117159
2.086045378 - 1.149844645 - 1.835460438

only peaks 2 shows stability.

next i looked how it would theoretically look if the injections were the same concentration (area of peak/conc)

peak 2: - peak 1: - combined areas:

262826 - 1035436 - 1298262
274360 - 208154 - 482513
279846 - 161158 - 441004
306437 - 195894 - 502332
285310 - 104278 - 389588
297585 - 59952 - 357536

again only peak 2 is close to something.
so i think that in this case we need to take another aproach concerning this problem. for them to be isomers we should have had the stable result with the combined areas and not with any one of the individual peaks.


Olivier do you have enough data to do the same number check?
again if your peaks are isomers the stable data (RSD, reproducibility of area) will be with the sum of the areas and not any individual peak.

[unmgvar]

BTW, results that could indicate isomers would be:

both peak ares have good RSD or stable results and so the group area.
none of the 2 peaks have good RSD or stable results, but the group area does.

[juddc]

amaryl,
for them to be isomers we should have had the stable result with the combined areas and not with any one of the individual peaks.

Not *always* the case. If the two isomers have markedly different spectra, then you could see significant variations in total peak area. For instance, molecules that undergo a keto-enol tautomerism can have markedly different spectra between the two species. The diketo form of avobenzone (butylmethoxy dibenzoyl methane) has a lambda max at ca 267 nm and the enol form has its lambda max at 360 nm. If the chromatography conditions don't strongly favor one form over the other, then you can see some odd behavior. With that said, I don't know the structure of the compound the questioner is working on, so I could be utterly fulla beenz, but I think you have to be careful when applying such calculations.

[juddc]

Relatively recently published paper on analysis of this molecule. A quick skim reveals that their chromatography seems OK, but their choice of buffer for the pH they used is a bit odd I think (phosphate for pH 3.5?).

Still, here it is:

http://makeashorterlink.com/?C5552268C

[syx]

Forget about Cis-isomer. USP29 have told us that there are 3 main impurities other than Cis-isomer, i.e. Glimepiride sulphonamide, glimepiride-urethane, and glimepiride-3-isomer (meta-isomer?).
Determination of Cis-isomer is individually confirmed using normal phase chromatography. I am deeply in doubt that Amaryl’s chromatogram showed this isomer.

[unmgvar]

Juddc,

your remark is correct.
i didn't think about it. i assumed that it would not be the case. in the last lab i worked for they had 2 aplications with isomers that had no changes in the chromophore site and so the analyses were done by simply summing the areas.
but anyway as Amaryl's number show, peak area 2 is "stable". peak area 1 behave in a strange manner. if it were an isomer case we would be having some sort of weird changes on peak 2 as well as on peak 1, and we would see it in the data.
there is something else going on here for sure.

[amaryl]

Forget about Cis-isomer. USP29 have told us that there are 3 main impurities other than Cis-isomer, i.e. Glimepiride sulphonamide, glimepiride-urethane, and glimepiride-3-isomer (meta-isomer?).
Determination of Cis-isomer is individually confirmed using normal phase chromatography. I am deeply in doubt that Amaryl’s chromatogram showed this isomer.

Thanks to all of you... so much discussion on haunted peak!

Siswanto, if you are referring that article over isomer determination in crude samples of glimepiride, the one I send to you, thats RP HPLC.

And if, its something else can you mail me that article.

Thanks and regards,

Amaryl.

Siswanto, if its not Cis-isomer, its something to have existence in chromatogram !

[Uwe Neue]

Syx, what do they call the cis and trans isomers. At what point in the molecule is the cis-trans transformation? I am used to cis-trans around a double bond, but I can't see it here?

[Oliver]

Wow i was absent for some time and this post has grown quite a lot.
the compound whose chromatograms i posted in the very begining was Amlodipine Besylate but we have observed this with Ramipril. With the latter we tried extending the run but no other peak was observed. I should still have the solution and would like to run it again. Strange thing when new standard was made this behaviour stopped.

There were no abnormal temperatures or temperature fluctuation in the sample rack as it is closed, besides the room is airconditioned.

Uwe: which book, I've read the one about the columns and I have the waters trobleshooting guide (yellow paperback A4 size)? I can't remember anything regarding this peak splitting.

unmgvar: If i'm not mixing things up i think we did do a sum of the areas but they did not come to a similar value for all the runs. If i remember correctly there was quite some variability. I'll try to check it again this week.

[Oliver]

As I'm reading the posts about isomerisation, i would still query why this happened now? There was nothing different AFAIK this time, and a new preparation did not show these sympthoms. I don't know if it would be possible that a contamination might have triggered the slow isomerisation process....

[Uwe Neue]

Oliver, my book on "HPLC Columns" has a troubleshooting section at the end. In this troubleshooting section, I show examples of these double peaks. The ...prils are all known to do this. With one of the ...prils, I got a dimer, which gave me a triple peak.

However, your compound, the glimepiride, is not a ...pril, but it has some similarity around the 5-ring. This is where the idea came from that you may be dealing with the same event.

I also have Troubleshooting Column that deals with this, but it may have been published later than your version of the Troubleshooting collection. I will send you a copy of this article, if you give me your e-mail.