Problem with C18 column on preservatives

[vincent001]

Hello everyone!

I work for a very very small company wich distributes VARIAN consummable.
Although I'm not a specialist in chromatography(studied a long time ago ), I try to help one of our customers which has a problem with his chromatogram.

He works in a cosmetic company and needs to establish a procedure for the quantification of preservative in his products.
For now, he's working on conservative (especially parabens) as a raw materiel, so its pure.

He had a problem with his former column (a VARIAN Microsorb 5µm C18 4.6x150 mm), which disapeared after using a new column Microsorb 3µm C18 4.6x100mm) : the peaks became wider, and he didn't manage to separate n- and iso- butyl paraben

Recently, the same problem appeared again, with both columns (he reinstalled his first column and chromatograms were OK). He sent me a chromatogram from his results:
He injected the same sample for 3 days, without changing the mobile phase, and without using any buffer.
Each color reperesents the chromatogramm of the day (during 3 days :day1 black, day2 red, day3 blue)




Focus on the n- and iso- butyl paraben



The method he uses is :

Column Microsorb MW 100-3 C18 100x4.6mm, oven temperature=40°C, Flow= 1 ml/min

0 to 6' : H2O/MeOH 60 :40

6 to 10':H2O/MeOH 60 :40 to 40 :60

10 to 18' :H2O/MeOH 40 :60

18 to 19' :H2O/MeOH 40 :60 to 60 :40


Is it normal that the 3 chromatograms showed different retention times, and do you have any idea why the resolution between n and iso butyl paraben decreased?

Thank you very much in advance for you help!

Bye

PS: sorry but my english is not perfect

[sassman]

It doesn't say if he re-equilibrates the column to the initial mobile phase conditions after each run. If not, it could cause shifting retention times especially during the first few runs. I would suggest re-equilibrating at H2O/MeOH 60 :40 for at least 3 minutes after each run.

[juddc]

It could be an equilibration issue...

Is he sure that his system is mixing the mobile phase properly / consistently?

Is the pump running a constant flow rate?

Cosmetics can certainly alter a column's performance. These samples may be "pure", but what else is the column being used for? I might advise a thorough washing of the column with neat ACN or even something stronger, then reinject once he's re-equilibrated thoroughly. In my experience, 3uM columns will foul a bit more quickly than their equivalent 5uM brethren without the use of careful cleanup procedures / guard columns, etc...

[tom jupille]

A couple of things:

1. The retention time shift is only a little over 1% (about 0.2 minutes). A slight variation in mixing could cause that (as could a slight flow rate error or small equilibration differences). Retention times on the "day 2" chromatogram are actually a touch shorter than on "day 1". Is your customer seen a continuous change, or random variations?

2. The retention time difference between your two peaks is virtually identical on all three days. The decrease in resolution is due entirely to broader peaks (fewer plates), and there it looks like a continuous decrease (day 3 < day 2 < day 1).

The problem could be either sample-related (more likely) or mobile phase-related (less likely). One way to tell is to run a series of injections of the same sample, pause for a while (keep running "dummy" gradients with no injection) and then run another set of injections. Then plot resolution versus number of injections and again versus elapsed time. If the problem is sample-related, the plot versus number of injections should be reasonably linear, but the plot versus elapsed time should show a "shelf" during the "no injections" period. If the problem is mobile phase-related, then the plot versus elapsed time should be reasonably linear, but the plot versus number of injections should show an abrupt drop between sets.

[Consumer Products Guy]

I'd concentrate on the column re-equilibration, and add a little acetic acid or phosphoric acid to the mobile phase, that's what we do. However, we also do this isocratic mode on RP-18 columns.

[vincent001]

Thank you very much for all your prompt answers!

I'll see tomorrow with him to investigate if it's either a sample or a phase problem.
Anyway I'll try to have more information about the way he reequilibrates his column as well as possible variations in the mobile phase composition.

>tom jupille : I definitely agree with you on the fact that peaks are getting broader in time, I'll ask if its always like that...

Thanks again!

[vincent001]

Hi!

I spoke to my customer today, he told me that after each day, he equilibrated his column with 100% MeOH during 40 minutes.
The day after, he ran a 60/40 H20/MeOH during at least 10 minutes to reequilibrate his column.

An important point is that the drift he observed with his peaks getting broader
only occured between two days, but the peaks were similar during the same day.

Now he had another problem, after he tested his method with AcN (which didn't allow him to separate the butyl parabens), and now that he get back to MeOH/H2O, he observes shoulders on all his peaks.

Do you think all these problems could be due to head spaces, bubbles or anything else?

Thank you

[WK]

Vincent
I get problems with retention times if I only re-equilibrate for only 10minutes - I use at least 20 minutes at 1ml/min.
I would also leave at least 7 minutes at initial conditions before starting next sample.
WK

[HW Mueller]

Could there be a confusion between re-equilibration to starting conditions after each run with the shutdown of a column over night?

[tom jupille]

An important point is that the drift he observed with his peaks getting broader only occured between two days, but the peaks were similar during the same day.

That would seem to point the finger at overnight storage conditions, as HW implied.

he observes shoulders on all his peaks.

The same kind of problems on all the peaks usually indicates a physical cause (headspace or partially-plugged frit), but I'm puzzled as to how/why a switch from MeOH to ACN would cause that. The problem would make sense if the column had been exposed to or stored under alkaline conditions.