Agilent GC - AUX

[Jana P]

Hello everyone

I was wondering if anyone can explain to me in a basic fashion what the AUX 3 (also AUX 4, 5) do in the Agilent GC?

I know that when I have AUX 3 set at say a pressure of 9 psi I get one result but if it is set higher, say 25, I will see a drop in the peak areas and it also appears to have an effect on the amount of analyte I detect in my samples.

I am new to GC method development and am seeking some guidance. I am trying to learn via hands on and a lot of reading but am struggling a little.

I appreciate any comments and advice.

Thanks in advance

[GasMan]

The AUX pressure controller is an electronic pressure controller that has three channels of pressure control. Depending on what other options are installed on your GC, up to three of these units can be installed on the 7890, giving 9 channels of pressure control.

These units always work with some other device on the GC, such as a rotary valve, an effluent splitter, a Deans switch, a backflush device, etc.
You will need to find out what your channels are connected to before we can explain what your particular channel is doing. If you can give me the serial number of the GC, I may be able to help.

Gasman

[Jana P]

Okay I am with you so far.

We have a 6890N GC System (Serial number US10650080) I am not sure what device the EPC unit is working with, I think it is a Backflush unit but am unsure.

I will try to find out for confirmation.

I appreciate your time.

[GasMan]

The 6890N delivered to you did not have any valves, splitters or back flush devices installed. A head space sampler was also part of the order and I think that probably one channel of your Aux pressure controller is connected to the head space sampler for the purpose of pressurizing the vials. This would explain why you get different answers when you change the pressure.

This is a case when you probably should read the head space manual and not the GC manual.

Gasman

[aldehyde]

Vial pressure default is around 14 PSI although I sometimes see lower or higher (like, 23 PSI). It is application specific.

What are you trying to analyze?

[Jana P]

Thanks for your input so far. Sorry I have not had a chance to reply sooner but I have been on holiday.

We see this AUX having an effect with all of our headspace methods. This to me now makes sense.

We see the most significant differences in peak areas and results (compared to alternate instruments) when trying to detect the residual ethanol in one of our products. Despite the method having been validated on both instruments (our Agilent and Perkin Elmer) and determined to be accurate, precise and linear. At the time of validation the results obtained on the two instruments was comparable (within acceptable ICH Guidelines for precision) using a real product sample. Since validation we have seen significant differences between instruments (1000 to 5000 ppm), mainly due to the Agilent providing variable results. the PE is consistent.

What is the best method to determine the appropriate vial pressure to be using (Setting the AUX 3 at)? Do I conduct accuracy studies and the correct vial pressure is when I am obtaining a recovery at 100%?

I am sorry if I seem to be misunderstanding but we have no GC analysts/experts on hand in our company or on site to offer any guidance with further development. It does not help that people want the problem fixed but don’t want to allow the time required to fix the problem.

Currently we run majority of our methods on our old PE GC/Headspace system and given it’s age I am fearful that in the near future I will be stuck in the middle of a mad rush to transfer methods to our newer Agilent’s when the PE finally kicks the bucket. I don’t want to be trying to resolve all the foreseen problems then.

I really appreciate the time you all take on this forum. sorry again for the late reply.

Thank you

[krickos]

Hi Jana

A not too easy question, but will have a go at it.

First I would dig up raw data from the validation and find out what vial pressure settings was used then. This is necessary for the validation status of the method/methods.
Once this is done you face 2 options: Either use the settings used in validation (if it seems OK) or change it and you will have to deal with revalidation issues.

Generally regarding vial pressure on Agilents headspace samplers: Well some call vial pressurization a source of errors and a diluter which is true in a sense. If your headspace oven temperature is high enough the pressure in the vial will be enough to perform a reproducible injection. In those cases one can still have the vial pressure on (200kPa) but set the pressurization time to 0,0 minutes, just to be sure that no sample can flow in the wrong direction.

The agilent headpspace operates with a pressure/loop injection which is different from older PE instruments such as HS-40 (balanced pressure). agilent system open ups the vial pressure against the atmosphere so basicly the vial pressure needs to be above atmosphere to get the gassample moving.
I can not claim that our "default" vial pressure when needed is entirely scientifly based but we tend to use 200kPa and keep it shorter in time (~0,5min) than typically on the PE instruments, and it works mostly for our purposes.

The vial pressure have impact on the "Loop fill/vent time" but normally I would say a setting of 0,2-0,4min would work.
The vial pressure is only one player in your headspace settings, what you want is to catch a reproducible "heartcut" of your gas sample and transfer it onto the column. In other words vial pressure may be ideal but still you can have issues.

For further insight in static headpspace I would recommend you to get hold of a copy of "static headspace-gas chromatpgraphy, theory and practise" by Bruno Kolb and Leslie S. Ettre; or something equivalent. So you at least can look things up.

And finally regarding a transfer from PE to Agilent. You might get away with a default setting on the agilent uinstrument, but you might aswell end up with doing individual tweaks especially if you have tight signal to noise demands on class 1 solvents for instance.

[Peter Apps]

Hi Jana

I presume that you have a G1888 headspace sampler ? I did some very intensive optimisation work on one of these, analysing aqueous ethanol at % levels. The instrument and its control via the GC have some "idiosyncracies" that you need to take into account. Perhaps the strangest of all is that when the pressurize time is set to zero the pressurize solenoid still cycles, so you get a very short (and non-reproducible) puff of pressurization into the vial. Also, if you monitor the vial pressure during pressurization you will see that it overshoots and fluctuates on its way to the set point (possibly dues to the long pipe connecting the AUX regulator to the headspacer). The temperature on the sample oven of the particular instrument that I had had about a 0.5K ripple, and fluctuated by as much as 2K when a room temperature vial dropped into it. In its default configuration the best repeatabililty that I could get was about 3% rsd on raw peak areas. With some drastic modifications I got this down to about 0,2%.

Since you probably do not want to re-engineer the hardware I suggest a vial pressurization that adds only a little gas above the pressure in the vial at whatever temperature you have the oven set. Put the pressurization time to about 30s so that whatever is happening has enough time to happen, be prepared to increase this time. The loop fills quite quickly, and you can monitor the flow of gas out through the vent at the back of the instrument to see when it has filled and equilibriated with atmosphere (there is an option that provides EPC control of the loop pressure, presumably by an AUX channel). The injection time should be longer rather than shorter - once the loop contents are in the carrier gas stream a longer time makes no difference, but cutting the loop out of the flow half way through will not be repeatable.

On the instrument that I had you could select on-board gas control that used mechanical regulators in the G1888. This has the advantage that the gas control is closer to what it is controlling.

How are you connecting the G1888 transfer line to the GC ?

Peter

[aldehyde]

I have absolutely seen some oddities when using the HS chemstation addon for the 1888. I have also seen pressure fluctuations with the AUX if the pressurization time is left at its default, I think I actually increased it just slightly but I would watch the vial pressure setpoint from the front panel during injection and make sure it isn't changing too quickly during the pressurization and equilibration just before injection.

Also, there is a new version of the headspace control software:

A.02.02 Headspace Software is Released!
A.02.01 headspace software has been revised to A.02.02.

Parts Required:
The new headspace software will be available in SubscribeNet. The CD part
number is G2924-60706.

Situation:
This revision of headspace software addresses these key issues:

1) A.02.02 is compatible with E.02.01 GCMS ChemStation.

2) A.02.02 supports the 7820 GC configuration.

3) G1888 error messages are better consolidated between the headspace
hardware and software.

4) Method recommendations for DMSO have been modified.
To access method recommendations, go to Headspace Method: Assist: USP
(DMSO).
Previous set points were: oven 160, loop 175, transfer line 185, Loop fill
time .2.
New temperature set points are: oven 85, loop 100, transfer line 120, Loop
fill time .02.

5) It was not possible to edit a running sequence on headspace software
revision A.02.01 with GC ChemStation. This has been fixed in A.02.02.

6) In A.02.01 and previous revisions of headspace software, the MHE default
parameter is "ON" with 2 injections. The default parameter should be
OFF. This has been fixed in revision A.02.02.

7) A.02.01 Headspace Software and B.04.02 SP1 GC ChemStation with RC .NET
driver selected in config invalidates communication with the 6850 and 6890
GC. This is fixed in revision A.02.02 headspace software.

Solution/Action:
The items fixed in this release and known defects can be viewed in the
headspace software SSB, available from the external Agilent.com site.

Additional Information:
• Headspace software is currently NOT compatible on Windows Vista.

• A.02.02 headspace software is compatible on Windows 7 32 bit.

• Headspace Software Revision A.02.02 does not work with the combination of
E.02.01 GCMS Software and 6890 Rapid Control driver. If a headspace is
configured to a 6890 instrument that is using the rapid control (RC) driver,
the GC will not start. The temporary solution is if A.02.02 headspace
software is installed with E.02.01 GCMS ChemStation and a 6890, the 6890 must
use the classic driver. If this is not an option, the headspace must be
operated stand alone. This is captured in the G2923AA SSB, KPR # 52.

Date: 8/11/2010

[Jana P]

Hi everyone

I just wanted to thank you all for taking the time to offer some advice and answer my simple questions.

I really appreciate everyone on this forum for taking the time to educate those of us with limited GC experience. I have found it all very helpful and informative.

Thanks again