UPLC and Agilent 1200

[chemist91]

I just want to know if someone have an idea of what is going on in this scenario. We are doing a method validation transfer. The other company did the method validation using an Agilent 1200 (RRLC), we are using an UPLC (Waters). The rest of the method is the same column, mobile phase, injection volume, gradient, etc...no changes at all. The only change is the system Agilent 1200 and Waters UPLC.
When we try to run the transfer here two peaks are missing in our UPLC...in their report the peaks are there. We run the sample in a HPLC and the peaks are there.
Any ideas why the peaks appear on the Agilent RRLC and not in the Waters UPLC. We are using the same column for both.

Thanks

[unmgvar]

2 first things come to my mind right now as a start:

is there a shift in the retention time? check the gradient delay volume.

what is the shape of your peaks? how wide are they? check the detector settings of both systems. any difference in the flow cell sizes between both systems?


can you get us chromatograms?

[Uwe Neue]

Two possibilities: gradient delay volume or temperature. Run a test of the gradient delay volume on both instruments (gradient with a small amount of acetone) and report the gradient delays. We can then recommend corrections.

[chemist91]

Thanks so much for your answers...

Retention Time and shape of the main peak is the same in both systems. Unfortunatelly we only have the UPLC and the original method was validated on an RRLC so we don't know about the flow cell sizes.

About the gradient delay I will think the retention time of my first peak also will change and will not compare to the RRLC and that is not the case.

Both chromatograms look the same, same peak shape, area of the main peak....but in the UPLC the two impurities are missing....we run the same sample in a HPLC and the Impurities are there.

Something has to be different bettween the UPLC and RRLC since we cannot see them in our UPLC.


How I run a gradient delay? but anyway I cannot compare since I don't have the RRLC..

Any more ideas?

[Uwe Neue]

The standard detector flow cell is larger on the RRLC compared to the same on the UPLC.

You argue that gradient delay can't be the problem because your first peak comes out at the same time. However, this argument assumes that your first peak sees the gradient, which may not be the case.

Add some gradient delay volume to the beginning of your gradient, i.e. run the gradient starting composition isocratically for half a minute, one minute etc, and see if the missing peaks reappear.

The other possibility that I did not consider before is that the missing peaks were some type of system peaks...

[danko]

chemist91,

Are these 2 impurity peaks very small? Is the main peak equally high in both systems?
Maybe you should inject a larger volume of the sample solution in order to dismiss potential sensitivity differences.

Best Regards

[chemist91]

The original method it was run in a RRLC and it wasn't run by us. We just have a copy of two chromatograms one of the sample in diluent and the other in placebo. With the sample in diluent we don't have any problems and both chromatograms (UPLC and RRLC) are the same, plus we can see the impurities.

I will try the gradient delay tomorrow to see if works.

In the sample in placebo the main peak seems to have the same peak area in both chromatograms, We made a big spike of the impurity in lacebo and still doesn't came out in our UPLC.

Now we did notice that if we dissolve the sample in diluent instead placebo then we are able to see the impurities at the same retention time in the uplc; but in the original method this sample is diluted in placebo and it is a chromatogram showing the two impurities. Also the baseline of the placebo it is lower so we can tell this sample was spike in placebo.

It is something in the placebo that react with the UPLC and not the RRLC? It is the same column...the only difference it is the system...there are other impurities and this will show up, however they are closer to the main peak.

[seaclimb2001]

The standard detector flow cell is larger on the RRLC compared to the same on the UPLC.

You argue that gradient delay can't be the problem because your first peak comes out at the same time. However, this argument assumes that your first peak sees the gradient, which may not be the case.

Add some gradient delay volume to the beginning of your gradient, i.e. run the gradient starting composition isocratically for half a minute, one minute etc, and see if the missing peaks reappear.

The other possibility that I did not consider before is that the missing peaks were some type of system peaks...

I have used both Waters and Agilent detectors and I can tell you this. The Waters detector, with its Taper cell, is much less sensitive to refractive index changes. Peaks near the solvent front that you may have been accustomed to seeing on the Agilent system, might not be there on the Waters if it turns out they were salt or solvent peaks.

[chemist91]

I will try few things today however I was already told that we will be using HPLC since we cannot see the peaks in the uPLC and we running out of time. So I guess I will no be able to try to make this work anymore.

We can see all the impurities in the HPLC, however the UPLC was more attractive because run time and solvent savings.

Thanks to all for your comments.