Perplexing HPLC issue - area of analyte peaks halved

[hplctroubleshooting]

I'm getting areas that are about half the value of what I'm used to , there's also slight variation in the areas , I've searched for leaks , replaced the lines , purged everything , and just about everything else ... there were two instances where the areas seemed to return to normal values and those were when the sample loop connection - inlet - to the autosampler was taken off and then put back on , made sure that it was screwed on pretty good afterwards and the area values returned to normal for about 2-3 runs .

I checked the flow of effluent on the sample loop by disconnecting the output of the sample loop and checking the flow , it seemed okay , the compression test passed , the only thing that I can think of is the valves but that doesn't really explain why things suddenly returned to normal while dealing with the sample loop ... repeated the disconnection/reconnection of the sample loop inlet and with no good results.

Online sources point to some kind of a detector pool as a source , but I have a waters hplc and it doesn't seem to have such a thing , something that's been confirmed with a waters technician. Anyways , are any of you familiar with what's going on here?

Help would be appreciated.

[tom jupille]

If *all* of the peaks on a chromatogram have the same symtoms, then it is very likely that the problem ocurred *before theay had started to separate (i.e., at or before the column inlet). That means you are looking in the right area.

I can speculate on syringe problems (aspirating air, perhaps?) or insufficient retention to get away from t0 noise?

[hplctroubleshooting]

If *all* of the peaks on a chromatogram have the same symtoms, then it is very likely that the problem ocurred *before theay had started to separate (i.e., at or before the column inlet). That means you are looking in the right area.

I can speculate on syringe problems (aspirating air, perhaps?) or insufficient retention to get away from t0 noise?

thanks for the reply , I replaced the syringe and checked that it didn't have air bubbles after emptying and filling ... I suspect a valve issue and not a leak since the retention times are the same as when it was running normally and there isn't any tailing or fronting however I just don't get why disconnecting and then reconnecting the sample loop inlet improved things , even momentarily. The absorbance is high enough to negate noise as a factor.

[tom jupille]

Despite what I said earlier, you can't ignore the detector. Especially a PDA, since they are typically "single beam" and can't compensate for fluctuations in lamp output.

At this point, it might be best to start from square one. Hopefully, you have records on how the system performed the last time Performance Qualification was done. If this were my problem, I'd go back and re-run PQ on the detector (wavelength accuracy, noise, and drift) and the autosampler (repeatability and linearity) and see what's different.

[hplctroubleshooting]

Despite what I said earlier, you can't ignore the detector. Especially a PDA, since they are typically "single beam" and can't compensate for fluctuations in lamp output.

At this point, it might be best to start from square one. Hopefully, you have records on how the system performed the last time Performance Qualification was done. If this were my problem, I'd go back and re-run PQ on the detector (wavelength accuracy, noise, and drift) and the autosampler (repeatability and linearity) and see what's different.

I appreciate your advice , repeatability and linearity are already off , I'll be checking on the detector to see if it's contributing towards any extent to the problem at hand.

[John Lim]

If there is a way to remove the autosampler from the system and replace with a manual injector to perform full loop injection tests into the HPLC then one can isolate autosampler from the rest of the flow path.

If that is not possible, then (re)consider other sources within the autosampler. You reported that you have already replaced the syringe, checked visually for air bubbles and reconnected tubing/fittings.

Some other recommendations...(these are general autosampler recommendations):

* Injection port seal: has this been replaced? This is usually replaced during PM cycles and a worn injection port seal can often result in the symptoms you describe (lower peak area response, poor precision/injection linearity). Consult with your vendor's Operating Manual for the replacement operating cycle recommendation.

* Syringe: often there is a PTFE gasket in the syringe valve for the syringe port and in other syringe valve inlet/outlet ports that can be worn from repeated opening/tightening which will allow air entry into the injection path; if these have not been replaced then consider doing it (these are also replaced during PM cycles)

* Wash solution: as Tom indicated earlier, air bubbles are problematic; is your autosampler wash solution degassed before purging the transfer lines?

* Ausampler method parameters: even though you reported that your autosampler precision and linearity tests are poor, please double check your autosampler method parameters? Is the syringe volume correct? Injection volume OK? Injection method OK?

* Injection valve PM: You mention the injection valve as a possible source issue. You did not mention if you did the PM on the valve by rebuilding it by replacing the stator so this is another possible source if there is microchanneling from a worn valve.

If all the above check out then you will need to continue your troubleshooting downstream from the injection mechanism. I don't know if you are testing with a column inline...it would be best to replace the column with a backpressure loop as in an IQ test to eliminate the possible effects of the consumable (e.g., if there is a column capacity or other column performance related issue).

I agree with Tom that your absorbance detector performance (lamp energy, optical path issues?) may be the source of your problem. Please investigate the other recommendations first before going the hard route.

I hope these suggestions help...good luck!

[hplctroubleshooting]

If there is a way to remove the autosampler from the system and replace with a manual injector to perform full loop injection tests into the HPLC then one can isolate autosampler from the rest of the flow path.

If that is not possible, then (re)consider other sources within the autosampler. You reported that you have already replaced the syringe, checked visually for air bubbles and reconnected tubing/fittings.

Some other recommendations...(these are general autosampler recommendations):

* Injection port seal: has this been replaced? This is usually replaced during PM cycles and a worn injection port seal can often result in the symptoms you describe (lower peak area response, poor precision/injection linearity). Consult with your vendor's Operating Manual for the replacement operating cycle recommendation.

* Syringe: often there is a PTFE gasket in the syringe valve for the syringe port and in other syringe valve inlet/outlet ports that can be worn from repeated opening/tightening which will allow air entry into the injection path; if these have not been replaced then consider doing it (these are also replaced during PM cycles)

* Wash solution: as Tom indicated earlier, air bubbles are problematic; is your autosampler wash solution degassed before purging the transfer lines?

* Ausampler method parameters: even though you reported that your autosampler precision and linearity tests are poor, please double check your autosampler method parameters? Is the syringe volume correct? Injection volume OK? Injection method OK?

* Injection valve PM: You mention the injection valve as a possible source issue. You did not mention if you did the PM on the valve by rebuilding it by replacing the stator so this is another possible source if there is microchanneling from a worn valve.

If all the above check out then you will need to continue your troubleshooting downstream from the injection mechanism. I don't know if you are testing with a column inline...it would be best to replace the column with a backpressure loop as in an IQ test to eliminate the possible effects of the consumable (e.g., if there is a column capacity or other column performance related issue).

I agree with Tom that your absorbance detector performance (lamp energy, optical path issues?) may be the source of your problem. Please investigate the other recommendations first before going the hard route.

I hope these suggestions help...good luck!

very helpful advice , much appreciated. we had a waters technician come in and did some auto runs after filling the vials with water , the vials were still full even after the injection , in fact , the caps were wet which indicated that the injection valve wasn't diverting the mobile flow while it was making its injections thereby spilling mobile phase onto the top of the vial caps. He replaced the injection valve in its entirety which costs us 3000 dollars just for the parts , and then replaced the filter at the outlet of the tubing that leads from the syringe to the system flow. It worked for a day and now the problem is evident once again.

I've never actually opened up the injection valve myself , is a lot of technical expertise required to replace the stator on the injection valve?

[John Lim]

To answer your question,

"I've never actually opened up the injection valve myself , is a lot of technical expertise required to replace the stator on the injection valve?"

It is a fairly straightforward procedure to replace the injection valve stator (there might be O-rings and seals involved too). In many cases, customers often perform this type of routine maintenance in-house and can purchase PM kits to rebuild the injection valve from the instrument vendor. If one has never done this before, I recommend asking your service engineer how to perform this or attend a Preventive Maintenance class that most instrument vendors offer for their products. Of course, if not replaced correctly or internal components are accidentally scratched, leaks and other issues may be the result. In many labs, the customers purchase service contracts that include scheduled PM visits to avoid the hassle, especially if they are in a GxP environment.

In your case, since the problem appeared so soon after the replacement of the entire valve, you should consider contacting the vendor to repair this under warranty. If you are under a service contract then you are covered; if you are on paid day service they might still honor the service.

[hplctroubleshooting]

To answer your question,

"I've never actually opened up the injection valve myself , is a lot of technical expertise required to replace the stator on the injection valve?"

It is a fairly straightforward procedure to replace the injection valve stator (there might be O-rings and seals involved too). In many cases, customers often perform this type of routine maintenance in-house and can purchase PM kits to rebuild the injection valve from the instrument vendor. If one has never done this before, I recommend asking your service engineer how to perform this or attend a Preventive Maintenance class that most instrument vendors offer for their products. Of course, if not replaced correctly or internal components are accidentally scratched, leaks and other issues may be the result. In many labs, the customers purchase service contracts that include scheduled PM visits to avoid the hassle, especially if they are in a GxP environment.

In your case, since the problem appeared so soon after the replacement of the entire valve, you should consider contacting the vendor to repair this under warranty. If you are under a service contract then you are covered; if you are on paid day service they might still honor the service.

Thanks for all of your advice , it was really helpful.

[Chromatographer2010]

Going back to the first symptom with disconnecting the sample loop increased peak response. It may just be a clogged sample loop.

Also since the maintenance do you still get liquid dripping on top of the vials?

Also what kind of Waters injector are you using? Some of them don't even have rotor seals.