EPA 8260 Optimization (GC/MS w/ Purge-Trap & Autosampler

[Caged Alpinist]

Hello,

I have searched the forums and archives, so please forgive me if this has already been posted, or if it's too basic for this forum...

I am running EPA Method 8260 on:

HP (Agilent) 6890 GC with 5973 MSD

Samples are introduced via an OI 4560 Sample Concentrator (K Trap) that is fed by an OI 4550 Autosampler.

Column is a DB-VRX 20m x 0.180mm ID x 1.00um film

Front inlet: temp- 240, pressure- 19.5, flow- 0.3

Ramp: 9deg from 35-260

I recently inherited this instrument set in a state of disrepair. It has taken about 3 weeks to eliminate multiple problems in the 3 instruments, but now they are humming symphonically and I have passed my 3rd party proficiency samples.

My problem is that my run time is 1 hour 20 minutes with a 21 minute GC run. To me this seems insanely long for this column, but I am still having problems with some of the early peaks co-eluting.

Any advice on where to start optimizing this system (hardware? temps? flows?) to achieve shorter run times and better separation would be greatly appreciated.

Thanks in advance,

A.W.

[AICMM]

Caged Alpinist,

Very interesting nom de guerre by the way.... Can you post a chromatogram for us? Also, can you tell us what the total flow through the system is, that is split flow plus column flow? Also, are the ions the same on the early co-elutions or not?

Best regards.

[Caged Alpinist]

Alas, I am an Alpinist in a cage of rednecks and flatlands... But I am plotting my escape very soon...

Chromatogram from current method:



The ions are unique for the early eluters.

The ramp for this graph is 10deg C per minute.

So far, I have made the following adjustments:

Reduced purge time from 11.00 mins to 8.00 mins.

Reduced bake time from 25.00 mins (!!!) to 10.00 mins.

Increased desorb time to 6.5 mins from 6 minutes.

Reduced column flow from 0.8 to 0.6 (total system flow is now around 33.5)

I also made some adjustments to decrease the ramp increment during the temperatures with multiple co-eluting compounds -- with limited success in improving resolution. I am planning on trying to reduce it a bit further.

Currently with these adjustments the last compounds are coming off at 18mins... Run time is reduced by about 15 minutes overall, separation is somewhat improved but still poor.

Many Thanks,

A.W.

[krickos]

Hi

Found this high speed VOC analysis (EPA 8260, DB-VRX same dimensions, Tekmar P&T) at Agilent, can be found in their catalogue as well:

http://www.chem.agilent.com/Library/chr ... ms/372.pdf

[Caged Alpinist]

Hi

Found this high speed VOC analysis (EPA 8260, DB-VRX same dimensions, Tekmar P&T) at Agilent, can be found in their catalogue as well:

http://www.chem.agilent.com/Library/chr ... ms/372.pdf

Great! I'll give it a shot and see how it goes.

Thank you,

A.W.

[gpronger]

A bit of clarification, if you're GC run time is 23 minutes, with a cycle time of more than an hour, what is the system doing in between?

Typically, I am use to the P&T purging the next sample while the GC is finishing the prior run and cooling down. I will typically adjust the bake time so that the P&T is coming to ambient about the same time as the GC.

Greg

[Caged Alpinist]

Greg,

Part of the problem was that the P/T was set to a 25.00 minute bake and the auto-sampler system was not performing it's intermediate operations until the P/T was to a certain point. I know I have to watch for carry-over on the trap, but 25 minutes bake time seems excessive.

You have a good point about the P/T being timed to purge when the GC is still running, but I'm not sure if that's possible with the OI system. I'll look into that today.

Thanks for the input!

Cheers,

A.W.

[gpronger]

Caged,

Check literature on the trap, most are cleaned out in 4 - 7 minutes.

Greg

[AICMM]

Caged Alpinist,

If your predecessors were baking the trap for 25 min at a time, it is possible they killed it. Might be worth investing in a new trap at this point. 8-10 minutes is usually plenty of bake. Furthermore, 6 minutes of desorb is too much. Everything is pretty well off in 2-4 minutes with a personal preference for 2 minutes but your regulator might argue 4. I would try that even before replacing the trap.... Longer than that and you are just stretching out the CO2 and the water.

You have decent peak shape up high and yech down low so I suspect something low level being coughed up by your trap. Possible to post a spectrum or two? Considering your sensitivity, I would also increase your split ratio a bit while maintaining the column flow rate you currently have.

Best regards.

[Caged Alpinist]

Thanks for all the great advice. I have made a lot of changes, mostly with the purge/trap system, and some with the GC (Split ratio, ramp temps). I am running some standards and blanks currently to see how it looks. I'll post some chromatograms tomorrow...

Thanks again,

A.W.

[Caged Alpinist]

In case anyone else needs 8260 improvement references... I found this from Agilent as well:

http://www.chem.agilent.com/Library/app ... 0603EN.pdf

Regards,

A.W.

[Caged Alpinist]

Chromatogram with High Speed VOC Method linked above.

1st half of compounds have terrible separation. I am replacing the trap on Monday and will see if that straightens it out any.

[Ron]

If the purge and trap is taking so long to cycle then there is something wrong wih the purge and trap or the method set in the purge and trap. I am not sure what a K trap is, OI uses numbers and not letters for their traps. I suspect that the method is copied from a Tekmar purge and trap, and if so the parameters are set wrong. I would start by making sure the trap temperature during purge is set to a lower value than the trap can reach. On a Tekmar the trap has to reach the setpoint before the purge and trap comes ready. On an OI the rate of cooling is monitored, and if the temperature starts stabilizing then the instrument comes ready. If you set a temperature slightly above ambient like is done on a Tekmar the temperature tends to oscillate as the system alternates cooling and heating to reach the setpoint. Try setting temp during purge to 20C and see what happens.

[Ron]

Two things I forgot to mention, OI purge and trap systems usually do better with a number 10 trap, and don't do too well with VOCARB traps.

Desorb preheat improves peak shapes on the early eluters on Tekmar units, but tends to broaden early eluting peaks on an OI unit.

[Caged Alpinist]

Ron,

I am calling it a K trap I guess because it says "VOCARB K" on the boxes my traps come in. What kind of problems do the Vocarbs usually give to OI P/T's?

I think you are right on target about the temps being copied from a method designed for a Tekmar P/T. I am still waiting on my new trap to come in, as soon as it is installed and conditioned, I will make these adjustments and see how the separation looks...

Thanks,

A.W.