CARRY-OVER with column switching

[konkous]

Hello,

I'm trying to develop a method for the quantitative determination of a drug and its hydroxy metabolite by using on-line SPE with column-switching. The strange thing is that there is an extremely large carry-over almost equal to my middle calibrator even when i perform repeated injections of blank solvent. I've tried removing my HLB columns and replaced them with simple connections but the carry-over remained. Switching is performed with a 10-port valve which is connected to 6-port valve onto which the injections are performed. The next step was to connect the tubing leading to the mass spectrometer directly to the 6-port valve leaving the 10-port off. There was no-carry-over! My conclusion is that i have a huge contamination problem with the 10-port switching valve. Does anybody know a way i can solve this apart from replacing the whole valve? Thank you..

[MG]

You can take apart the valve and replace the rotor seal and/or stator face. Usually the manufacturer of the valve will sell rebuild kits, or you can get them from a third party such as Alltech.

[konkous]

Thank you for all the help. I have performed several additional experiments which show that probably the 10-port switching valve isn't the main cause of the problem. I replaced the HLB columns with 2 connections and the carry-over was almost eliminated. My next move was to place only one HLB column and leave one conection. When the analysis was performed on the HLB a huge signal appeared for all my compounds. On the contrary there was no signal when the analysis took place on the connection. I have to inform you that after the HLB's there is also an analytical column. So it seems now that the HLB's have the problem. The funny thing is that i keep injecting blanks and i receive the same signal intensity each time.

[Uwe Neue]

This indicates that your elution from the HLB column is incomplete. What are the details of the extraction and - more importantly - the elution protocol from the HLB column? Are you using a backflush to elute onto the analytical column?

[konkous]

I have run the specific method before and it is the result of thorough optimization concerning the carry-over problem. After all i have checked the HLB columns by passing my sample only through them without using the 10-port valve. There was no carry-over. Isn't it funny that each time i run a blank sample i get the same signal for all three compounds? Shouldn't it be smaller at each injection? After all, 25 blank injections after a system suitability test sample are more than enough. Another funny thing is that when i remove the HLB columns and replace them with connections there is no carry-over either were i to assume that the problem was my 10-port valve. I have also replaced all tubings on it..

[MG]

If it doesn't get any smaller with repeat injections, then perhaps a mobile phase or SPE loading solvent is contaminated with these compounds.

[Uwe Neue]

It is difficult to understand what could be issue without seeing the system and the way it switches the flow paths. I have seen strange things in the past, where sample got trapped in loops of the flow path and eluted in subsequent injections. If I were in your lab with you, this is where I would look.

[HW Mueller]

What do you get if you don´t inject anything, just running through your whole sequence?

[konkous]

I want to thank you all for your valuable help. It seems that the problem has nearly been solved. I washed all of my solvent bottles and filled them with new solvents. That did not seem to eliminate the signal which remained at the same levels for all of my three compounds. The next step i followed was to replace the valve inlet frits of both of my binary 1100 Agilent pumps. The signal was reduced by 20 times, at a level about half of my first calibrator. I don't know whether i should try cleaning or replacing the rotor seal of my 10-port valve.