GC response changes for PCB analysis

[josebenjamin]

Dear Friends,

I hope you can help me with the following situation; After a recent preventive maintenance my GC separations of PCBs (EPA 8082 method) have decreased 10x in response (875mV vs 76.1mV aproximately, measured for the largest peak). The retention has not changed and the makeup gas or detector gases have been kept the same. Column and flowrates are also the same. No leaks have been detected.

My basic question is how can a change in injector insert, resetting of the column connections, and a change of some gas filters, produce such a drop in response?. Are there any parts of the Varian CP 3380 model (injector 1177) so critical that could produce the observed change?

Perhaps you have observed similar problems in other analyses. I would like to hear your suggestions and experiences in dealing with these situations.

Thanks in advance

josebenjamin

[Peter Apps]

Whenever a signal drops (or increases) by a power of ten you need to check a range or attenuation setting on the signal output.

Peter

[sassman]

Aside from something like a change in attenuation setting or split ratio, my guess would be something wrong in the injection port. Is the injection liner exactly the same kind as what was installed before? How confident are you in your column cutting technique? Did you use an electronic leak detector to sniff for leaks?

[Don_Hilton]

Like sassman, the inlet jumps out to me also. I have seen things like a cracked inlet liner spoiling sensitivity. I would suggest opening the inlet, pulling the liner and begin sure that everything is in good shape and properly installed.

[josebenjamin]

Thanks very much for your comments. Most of the suggested checks and tests mentioned in your replies were done before the posting of my first message.

We will continue looking and testing no matter how remote the cause may look to be.

josebenjamin

[AICMM]

josebenjamin,

Three things come to mind immediately. First, column position relative to the gooseneck if you are using a constricted liner. The position of the column is critical in this case and only very important if you are using a straight wall liner. Second, are you using wool or were you using wool and do you have the same amount in the liner. For PCB analysis I would use wool. Third, is it possible that someone has switched you to a split injection when you were formerly doing splitless?

Best regards.

[minhtruc]

In general, lost in sensitivity is due to bad injection, bad instaled column. You are using ECD, right? ECD is concentration dependant detector which means that its signal depends on the make up gas. Have you tried to adjusted make up gas flow ?