Moving peaks

[gurkishan]

Hi Guys!
I am working with a mobile phase of 60% water, 39.9% methanol and 0.1% Glacial acetic acid.
I get good resoultion of 4 peaks that i want to start with but after some injections, the peaks start moving in.
Like my one peak comes at 6.2 min and after 10 injections, it reaches 5.5 min.
Pressure is constant, and i have changed new columns 3 times.
Cant figure out what is going on.
Any advice?

[Peggsy]

Are you running at ambient temperature and uncontrolled? RT's vary significantly with temp changes, as the temp increases during the day, RT can become shorter and vice versa for night runs. If this is the case, try and hold the lab temp more constant or better still, use a column heater/oven to maintain constant temp +/- 2C.


Other points to consider would be:
is the sample dissolved in the same polarity strength as the mobile?
is the separation isocratic?
check for column overloading.
does the pressure during the run stay constant?



Let us know how you go.


Pegs

[juddc]

This can also happen if you have something in the sample that is sticking to the column. Try rinsing the column off well with neat methano, re-equilibrating,l then reinjecting your sample. If the RT's go back to where they started immediately after you rinse, you probably have a contamination issue and you might need to put a column rinse step into your method.

[gurkishan]

Hi,
Thanks for the replies!
I am running in a column heater at 45C.
Mobile phase is water 60% methanol 39.9% and Glacial acetic acid 0.1%
So, no buffer, thus no sticking in column.
Also, i have tried 3 new columns, yet it is happening in all of them.
I ma giving the machine 3-4 hrs stabilisation time yet, peaks are moving-- sometimes inwards and then outwards, ie more like pendulum motion!

Will love more advice!
Thanks

[Consumer Products Guy]

(1) Are you pre-mixing your mobile phase, or letting a mixing valve mix the organic and aqueous components? You seem to state isocratic run.

(2) Are you purging ALL channels, even unused ones?

(3) Is your pressure steady throughout, or fluctuating? What is your column pressure?

(4) Is your vacuum degasser on, or have you solvents been degassed?

(5) Pump column eluent for twenty minutes into a graduated cylinder, see if your flow rate is correct (likely is, but easy to check)

If you are pre-mixing phases (1), I'd pre-mix a batch and put all channel leads in that reservoir, purge all the lines, to isolate whether a mixing vlave is causing this. Answer the questions and report back.

[gurkishan]

Yes,
I mix the solvents and sonnicate for 1 hour to deggas them
Pressure is constant throughout and is stable at aound 2252 psi with column heater at 45C.
I do purging of system before injecting.
Check valves are ok, flow rate from discharge has been calibrated.
MY system seems ok because i recently ran a different study which gave me awesome results but when ever i use this mobile phase, it gives me changing retention times.

[juddc]

So, no buffer, thus no sticking in column.

Not the case, necessarily. If there is an excipient in your sample that is extracted in the prep and it is sticking to the column, you can have retention issues. I've seen this most with mobile phases containing very little organic modifier (unlike yours) and/or with smaller columns.

What are the pK's of your analyte(s)?

[gurkishan]

hi,
i am basically running assay for aspirin, paracetamol and caffeine tablets.
I am following the USP monogram assay but it is a very bad method.
I talked to USP guys, they agreed it was a bad method but had no alternative.
my mobile phase pH is around 2 and that destroys the column very fast.
To prevent that, i decreased the glacial acetic from 3% to 0.1% to get pH of 3.5
Problem is, although i am having good resolution, but after 30 injections, all retention times vary with ~4% standard deviation.
No one is able to understand that why are the peaks showing pendulum motion in and out.

[Bruce Hamilton]

Frankly, you aren't " basically following " the USP procedure when you reduce a mobile phase component from 3% to 0.1%.

Follow the proscribed method, it's worked for many years, so if it doesn't work for you, try analysing a competitor's products. If you get OK results, look at your formulation.

I'd also be curious why USP staff would say the method is bad, as it's been used for years. If it was bad they should have fixed it.

Bruce Hamilton

[gurkishan]

i did follow the method for some time but my coulmn was getting destroyed in 3 weeks.
USP said that this was the prob, but they had no better method.
They adviced to follow something different and validate it and let them know.
But, as of now, i am facing this moving retention time prob!

[Bruce Hamilton]

I'm rather dubious that such a method would destroy suitable columns, even at 45C, but I don't work for a column manufacturer.

What is the symptom of the column failure?. Are you sure it's not attributable to sample or sample preparation? .

The USP should be able to point to column brands that have been successfully used for the assay, even just for validation.

Have you tried different reputable brands of columns, and/or talked to the column supplier?. I'm sure selling you a new column every three weeks would be good for business, but I would have approached them about the failure and asked if they had a more robusta nd appropriate columns. I'd also ask other major column manufacturers for suggestions.

I would have thought that validation of a new method to replace a compendial method was way more expensive than replacing columns, and also likely to trigger any client auditor's interest.

Bruce Hamilton

[gurkishan]

well, i totally agree with you!
But, we got columns from Waters, a reputable company, who, did replace 2 columns, as they felt that maybe their columns were bad.
I talked to tech support and they adviced me on calibrating the instrument, which i did.
Column going bad can be seen from high increased pressure and then tailing.
Cant think of much to do now, so hoping for some good advice

[Dan]

Perhaps this is a problem of acetaminophen (or excipient) build up on the column.

With a dosage strength of 500 mg acetaminophen per tablet and the limited aqueous solubility of acetaminophen, build up on column can occur. It can also be an excipient build up. We encountered this issue previously, see this link:

http://www.sepsci.com/chromforum/viewtopic.php?t=7837

We resolved the issue by using a wash solvent that had an organic content higher than that of the mobile phase. The wash solvent was used after every run (every 30-50 injections).

Regards,
Dan

[Bryan Evans]

A few options:

(1) Stick with the current column under method conditions (3% AcOH, 45C)
and live with the poor column lifetime (retention is probably stable at pH 2)

(2) Re-validate (or show equivalency) under the same experimental
conditions using an alternative column. Unison UK-C18 would be a great alternative
as the data shows this column is stable under low pH condtions.

Just curious - is caffeine retained at all under these conditions?

[Consumer Products Guy]

mobile phase pH is around 2 and that destroys the column very fast. To prevent that, i decreased the glacial acetic from 3% to 0.1% to get pH of 3.5

pKa of acetic acid wouldn't get a pH down to 2, especially at a 3% level; pH 3.5 seems more accurate. Are you measuring "pH" with the methanol added (if so, that's not really pH). And pH 2 is not so low anyway, we even use 8% H3PO4 in mobile phases with no issues.