retention time

[paolino]

Hello everybody.
I am now using a C8 silica column in 100% methanol as eluent.
I found that my substance peak retention time decreases from early in the morning (say 12.5 minutes) to the evening (less than 11 mins).
If I stop the flow and let the column rest for the night, the Rt gets back to the 'right' value.
If I don't, the Rt goes on shortening.
I haven't seen anything like this so far, although I'm not very skilled in NP columns. Does it seem like a hydrophilic collapse?
Thank you very much for your help.

[Srinivas SK]

Is this happening with all the peaks in your chromatogram? Check relative retention times of your peak of interest w.r.t other peaks. If relative retention times are steady while absolute RT's decrease, then it's probably due to the pump or due to the injector. Check the flow rate in that case. And check the injector for leaks.

If it's happening with one or two peaks only, then probably your sample is not stable under the conditions in use. In which case, prepare the sample fresh and try. Store the sample for a few hours and check again. If your sample isn't stable, you'll need to store it at sub-ambient temperature or use an amber vial.

Also, you can reverse the column and see if the same phenomenon occurs. If so, it's the sample that's causing the problem. If not, then you've got a column void, most likely. Change the column and check.

In case you're not stabilising the column before use, I'd suggest you flush the column with mobile phase for an hour at least before use.

Hope this helps.

[paolino]

1. yes, it happens with all the peaks. the pattern of the chromatogram is always the same and the relative retention times aren't varying very much.
I calculated RRts with two other peaks and the values change from 0.900 to 0.917 (earlier peak) and from 1.152 to 1.165 (latest peak).
Is it too much? I don't think so.
I've checked the system for leakage but I didn't find anything unusual.
I can't tell you anything about the pump itself by now.
2. I also checked that my substance is stable under HPLC conditions.
3. Firstly I thought it was an equilibration problem and I used to wait hours before starting analyses: but the Rts becomes shorter even afetr 12 hrs or longer.
4. as last chance I'll try to reverse the column

[Uwe Neue]

How is the temperature controlled? If you do not have temperature control of your LC instrument, this could be caused by changes in room temperature.

[paolino]

well, it is set at 25° C and controlled by a column termostatted oven. the oven is working properly.
when I had run this method on a longer column, retention times in the night were very longer. since this happened only at night, I realized that I needed a column temperature control, despite the fact room temperature in the lab was only four degrees lower at night than during the day.

[Uwe Neue]

we may still be on the right track.

How do you preheat the solvent before the column? Most column ovens have a loop included in the thermostat to preheat the solvent before it gets to the column.

The other thing is that many column heaters work only very marginally at such a small difference between room temperature and intended column temperature.

[amaryl]

Well even i experienced such situation. RT used to be quite stable at night.

Its probably room temperature and surrounding conditions. I agree with Uwe sir.

If their is no leakage, pumps working properly, no variation in mobile phase composition (pH molarity % B). Equilibration of column with mobile phase is even critical to attain stable RTs.

less chaos at night... HPLC works best

Plz even check sometimes allowing your pump to have wash for 5-10 mins ( opening drain valve) to release some mobile phase (air if by chance enters) even works and stabilises your RT. Air enrapment in this case will result in more pressure. Check pressure too.

Regards,

Amaryl.

A void in column will result in bad peak shape (broad tailed peak) even peak spliting and unusual peaks. I don't think reversing column will help.

[amaryl]

Don't reverse flush your column unless your pressure is too high (blocked column frit), unusual broad tailed peaks occur in analysis.

I don't think problem is too serious to subject your column reverse flush. Can effect packing.

Regards,

Amaryl.

[Srinivas SK]

From what I've read thus far, it's happening with all the peaks, RRT's are steady, temperature is controlled, sample is stable, column has been equilibrated before use and there are no leaks in the system.

Please check the flow rate accuracy. Depending on how old your pump is and depending on the drive system it uses, flow rates can vary over prolonged use. Cam-drive pumps can have fluctuations in flow rate, as they age. This is just a possibility! I don't know how old your pump is or if you regularly follow IQ/OQ/PQ procedures, but there's no harm in checking the flow rate anyway.

So I'd suggest you check the flow rate over the day. You may check with and without the column. It's easy to check and doesn't take much time.

If I'm wrong and your pump is OK, then breathe easy. It's the column most likely.

In which case, I'd suggest ( this is what I would do)

- Reverse the column. Disconnect from detector. Flush with mobile phase for about an hour ( to flush out the frit), reconnect detector, equilibrate system as usual, run your samples overnight.

The idea is to check what's happening, not to use your column long-term in the reverse position. Briefly reversing the column won't destroy it - in fact, it'll probably flush out the frit and reduce backpressure when you use the column in its normal mode.

If the same RT problem occurs, even on the reversed column, and if your pump is OK, then I must agree with Uwe - you need to take a hard look at the temperature control.

You can always check the temperature control by injecting a totally different sample on the same column. You can inject the test mixture that must have come with the column - or make one yourself.

You can also inject your sample of interest on a different C8 column or a C18 column, if you have one.

If RT's are changing as before, no matter which column or sample you run, then it is a temperature issue.

If it's happening only with one particular column and everything else is OK - pump, temperature, sample, mobile phase, etc.. then I'm afraid you have one very naughty C8 column.

Hope this helps. You can accept or reject all the above suggestions - but this what I would have done. There are no shortcuts I'm afraid.

Do let us know - one way or the other.

[amaryl]

From what I've read thus far, it's happening with all the peaks, RRT's are steady, temperature is controlled, sample is stable, column has been equilibrated before use and there are no leaks in the system.

Please check the flow rate accuracy. Depending on how old your pump is and depending on the drive system it uses, flow rates can vary over prolonged use. Cam-drive pumps can have fluctuations in flow rate, as they age. This is just a possibility! I don't know how old your pump is or if you regularly follow IQ/OQ/PQ procedures, but there's no harm in checking the flow rate anyway.

So I'd suggest you check the flow rate over the day. You may check with and without the column. It's easy to check and doesn't take much time.

If I'm wrong and your pump is OK, then breathe easy. It's the column most likely.

In which case, I'd suggest ( this is what I would do)

- Reverse the column. Disconnect from detector. Flush with mobile phase for about an hour ( to flush out the frit), reconnect detector, equilibrate system as usual, run your samples overnight.

The idea is to check what's happening, not to use your column long-term in the reverse position. Briefly reversing the column won't destroy it - in fact, it'll probably flush out the frit and reduce backpressure when you use the column in its normal mode.

If the same RT problem occurs, even on the reversed column, and if your pump is OK, then I must agree with Uwe - you need to take a hard look at the temperature control.

You can always check the temperature control by injecting a totally different sample on the same column. You can inject the test mixture that must have come with the column - or make one yourself.

You can also inject your sample of interest on a different C8 column or a C18 column, if you have one.

If RT's are changing as before, no matter which column or sample you run, then it is a temperature issue.

If it's happening only with one particular column and everything else is OK - pump, temperature, sample, mobile phase, etc.. then I'm afraid you have one very naughty C8 column.

Hope this helps. You can accept or reject all the above suggestions - but this what I would have done. There are no shortcuts I'm afraid.

Do let us know - one way or the other.

Take a deep breath. Problem is not that big. let her just state the pressure. If pressure is high quite possible blocked column frit which is always the last option to check why pressure is high.

RT variation is normal thing to happen with HPLC unless you have some good luck and you reproduce identical chromatographic conditions and surroundings.

I still don't support reverse flushing the column unless the pressure is too high and the peaks are distorted.

Regards,

Amaryl.

[amaryl]

Check pump, flow rate, flushing pump to release some air (by chance it enters), pressure, temperature.

Regards,

Amaryl.

[Uwe Neue]

Srinivas,

I generally do not recommend to reverse the column. For one thing there are some column that do not tolerate such procedures, for various reasons, from details of the packing process to the choice of frits. In addition, the problem described appears to be rather simple. Therefore especially in this case, I do not recommend it.

This does not mean that the column would not tolerate a reversal. It appears to be just an unjustified stress under the given conditions.

[amaryl]

Column is heart of HPLC Needs care!

Amaryl.

[paolino]

I thank you all for your help and advice.

I'm now checking the flow rate during the day. The pressure is very stable.
As soon as I finish, I'll have two different scenarios:
1. the pump can't work properly at that fixed flow rate
or
2. the column isn't the best column, a naughty C8.
In any case I think I'll have to review the method.

A further info is that now the Rts seem to be steady and my standard peak is now set at 10.7 mins, more or less. When the column was new, this Rt was 12.5 mins.
It seems that the column took two weeks to equilibrate...

I'll let you know the results of the flow check, as soon as it is over.

[Uwe Neue]

You never told us what your analytes are...