Strange peak splitting

[Oliver]

Hi, I am encountering a strange peak splitting problem and i have no idea where to start from. I believe this is mechanical problem but I don't know what could be causing it because most of the time the HPLC is working fine.

Below is an overlay of a sequence of chromatograms from the same vial done at the begining of a sequence of runs. This is a reverse phase chromatography, UV detector with + polarity (i.e. absorbance makes a positive peak). It seems that all of a sudden the peak split then recomposed itself after a couple of runs
The run sequence was first the black trace, then there are the 4 abnormal runs. They were first the dark purple, then the dark green, then the light blue, then the pink and finally it returned to the initial form, the green trace(congruent with the black trace).
Has anyone observed this behaviour

I have a similar problem from another HPLC, on an LC using an anion exchange column and nitric acid mobile phase giving negative peaks; that I will post in another posting.
Thanks for reading.

[unmgvar]

Hello Olivier,

from the behaviour you describe it is defenetly not a degradation.
have you thought that maybe your product is in a thermodynamic equilibration of two geometrical isomers?
looking at the peak height it looks like your 2 main peak of all chromatograms add to about 70 on your scale for all injections.
have you looked more closely at your other peaks in the chromatogram?
do they show also some distortion? generally when you have a system problem (hardware or column) it is not specific it affects all the compounds.

could the chemistry of your compound permit a ring formation for example or another change?
if yes then you will need to find the PH range at which the stability of your desired geometry is best kept.

or if of no consequences to you could always just go the easiest way and just show that both peak are isomers and simply add them together. there are some applications that work that way.

[amaryl]

This post is opening up the problem I had experienced. Same pattern i found working with glimepiride, standard for which was approaching expiry. Though I could not paste a chromatogram and the ghost peak still haunts !


I kept saying I think its isomers but my boss didn't agree to it.

My post is locked virtually and in reality !

Amaryl.

[jitender]

Not offering any solution to problem, but adding my own question.

That may be equilibration of isomers. But why the equlibration is so variable under same operating conditions. I understand HPLC is open system rather a closed one. Still any reason of such a variability in equilibration?

@amaryl- can you post overlapped chromatograms of your HPLC runs for same problem?

[amaryl]

These are the chromatograms for Glimepiride standard at concentration 0.195, 0.391, 0.781, 1.562, 3.125, 6.25, 12.5 ug/ml respectively.

The peak at retention time of 6.73 to 6.81 mins is of trans glimepiride (biologically active)...confirmed by PDA scan at 228 nm

The peak at retention time of 5.94 to 6.01 mins is the ghost peak which i doubt its Cis isomer.

Mobile phase 20 mM phosphate buffer (pH 3.0) : ACN (40:60 , v/v). Wavelength 228 nm. HPLC-PDA.


Regards,

Amaryl.


This time the pasting of chromatgrams worked. Thanks Tom sir.

[HW Mueller]

Something appears to be "fishy". All possibilities I could conjure up would have required highly unlikely coincidences, like:

If an equilibrium problem were behind these one would have to undulate some conditions drastically, but only ahead of the chromatography during which everything would have to be stable??

[Oliver]

Amazingly enough today i had another analsyt working on another product using another column and mobile phase, and we had the same problem. After lots of trials including extending the runs we came to no conclusion. In the end we decided to prepare a solution of a related active and inject that.... no peak splitting was observed!
Then we moved the sample rack to another HPLC set up in the same manner and we got the splitting again so i guess i can exclude the HPLC and blame the sample. We tried was loosening the vial caps to ensure no cavitation in the needle was occuring. At first this seemed to work but then we got the same problem after 4 runs.
Its late now so we stopped the trials and will restart everything from scratch on monday. However we kept the solutions so that if the problem does not manifest itself on a completely new setup i will inject these solutions and check if the phenomenon appears again.
Another thing worth nothing is that the new peak appears before the expected retention time so i am excluding double injection (i.e. injected portion going in the column in 2 stages)

[amaryl]

Something appears to be "fishy". All possibilities I could conjure up would have required highly unlikely coincidences, like:

If an equilibrium problem were behind these one would have to undulate some conditions drastically, but only ahead of the chromatography during which everything would have to be stable??

what equilibrium problem?

Help me this time!

Amaryl.

[Uwe Neue]

Do we agree that both peaks are glimepiride, as suggested by amaryl?

If we agree on this, the reason that we got two peaks could be related to the nitrogen in the 5-ring structure. This is similar to a proline, which is known to give slowly interconverting isomers around the nitrogen of the 5-ring. Actually, it could be much worse in this case, since you got carbonyls on both sides.

I have no clue though what triggered the variability from injection to inejction.

(actually, I might have some information on this somewhere...)

[unmgvar]

Amaryl,
can you performed repeated injections like Olivier at the same concentrations?
i believe that if the grouping of both peaks does give a good RSD then we can go and look up at this theory.
And/or with those injections you posted can you give us a linear curve of the group. if close enough to 1 then it is a possibility as well.

[amaryl]

Amaryl,
can you performed repeated injections like Olivier at the same concentrations?
i believe that if the grouping of both peaks does give a good RSD then we can go and look up at this theory.
And/or with those injections you posted can you give us a linear curve of the group. if close enough to 1 then it is a possibility as well.

I can't perform injections. I don't have access to HPLC now. I am working now with a pharmaceutical (not in analytical division). Well this had been my masters project where I experienced this problem.

Though I can provide all the data which I have as a soft copy. But working in a new way won't take place easily. Its tough out here in India.


Regards,

Amaryl.

[amaryl]

Conc 2nd peak 1st peak Combined
0.195 51251 201910 253161
0.391 97191 161105 258296
0.781 214275 162568 376843
1.562 437119 251729 688848
3.125 957617 612169 1569786
6.25 1783187 651735 2434922
12.5 3719809 749394 4469203


By group means combined area? plot of conc vs combined area regression coefficient = 0.9926
Plot of conc vs 2nd peak area regression coefficient = 0.9994

Thanks and regards,

Amaryl.

[syx]

Amaryl, there are other impurities than Cis-isomer in glimepiride (see USP29). Isomers would not be separated too easy like that. I do not believe that it is the Cis-isomer. Perhaps it was contamination from the glass containers.

[amaryl]

Amaryl, there are other impurities than Cis-isomer in glimepiride (see USP29). Isomers would not be separated too easy like that. I do not believe that it is the Cis-isomer. Perhaps it was contamination from the glass containers.

My orignal standard gave just one peak and with passage of time started exhibiting two peaks. Siswanto, you know it I checked at every possible cause, its not contamination from glass container. Even I got this ghost peak with glimepiride expired tablets and in degradation study (of large intensity). Though with fresh tablets it was small.

Siswanto, you have the reference of glimepiride isomer separation. They did on RP HPLC. I didn't find some extraordinary thing in chromatographic conditions to have such separation.

My guess may be even wrong it can be a impurity, a degradation product. Just wished to know whats the ghost peak that haunted me for long. Nothing more !

Amaryl.

[HW Mueller]

amaryl, I had assumed for a moment that the compound (Oliver´s) may isomerize with a, lets say, strong temperature dependency and that the temperature of the sample was reversibly varied in its container. No isomerization takes place during the chromatography, a bit far fetched?
Also: Your example just looks like it is going from one isomer to another while standing, but not during the LC. Nothing unusual if isomerization is slow in comparison to the LC.

Uwe, do you have a ref on this slow proline interconversion? This would be very surprising to me.