Setting up splitless method on Agilent 7890B GC

Hi John,

Great question!  I have a few follow-ups. 

What inlet liner is installed, and what is the makeup of your blank?  Is it pure ethanol?  Or ethanol in another solvent?  If this is a mixture, what is the concentration of Ethanol? 

Are you using an auto-sampler or performing a manual injection?  

What is your FID output?

I'll run through a few things to get hopefully help you out:

The optimal purge time can be estimated by dividing the liner volume by column flow. If you have a standard splitless liner installed (p/n 5183-4696, 900uL(0.9mL) volume), and we use your column flow of 1.8mL/min, the purge time value is 0.5 min.  If you are using a smaller volume liner your optimal purge time could be earlier.     I might consider a lesser purge flow (~20mL/min?) to conserve gas, but that will depend on your sample matrix and how volatile it is.  The purge flow will continue throughout the run unless you apply a gas saver setting, so with a 10 minute method, you would be using a liter of helium.  

Another value to be cognizant of is your total flow reading.  The total flow is a sum of the column flow value + Septum Purge flow + Split or Splitless purge flows, depending on mode.   Press Front or Back Inlet on the GC keypad(whichever applies to your inlet position), then use the arrow keys to scroll to the Total Flow line. 

In this case, column flow is 1.8mL/min, Septum purge is 3mL/min, and purge flow is 100mL/min.  Under these conditions, your total flow reading should be 104.8mL/min or very close.  

          If the value is higher than 104.8mL/min, there is a leak somewhere in the flow path, so you would need to evaluate the septum, column connection, and other inlet flow path connections as potential leak points.  The manual (link below) has good instruction for troubleshooting tests, or you can use an electronic leak detector to find the problem. 

          If the flow is reading lower, there is a restriction somewhere in the inlet flow path.  If you have a flow meter handy, measure the flow exiting the GC 'Split Vent' port on the top rear of the system.  The flow exiting the vent should be pretty close to your purge flow, ~100mL/min in this case.  

A plugged FID jet can also prevent samples from being detected.  Sometimes little shards of ferrule or column can enter the jet orifice and restrict flow. There is a jet restriction test built into the service mode/diagnostics menu on the keypad--just press service mode then find the FID diagnostics.  The fix is to replace the jet and trim then reinstall the column, preferable with a new ferrule.  These instructions are in the maintenance manual that should have accompanied your 7890B.  If you don't have access to the manuals, we do post them online.  The maintenance manual is here:  https://www.agilent.com/cs/library/usermanuals/Public/G3430-90052%207890B_Maintaining%20Guide.pdf

Are you seeing any changes at all on your FID signal?  You can verify the flame is lit by holding a shiny object (a wrench or mirror) over the detector and looking for condensation.  If the flame is lit, try changing your hydrogen gas flows and watching the output change. 

I hope this information is helpful to you.  Please keep us updated on your progress!

Abbey

Thank you Abbey for your response. I am still puzzled as to why I see no signal while in the splitless mode. I do obtain signals and chromatograms when operating in split mode.

The liner is a 5190-3162

Blank is ethanol (pure)

sampling via A/S

Been having some issues with FID output. This is a brand new instrument, I am using high purity gases and "Gas Clean" filters.

Have lowered the FID output threshold to 0.2 pA on suggestion of a field engineer - was having the flame go out when set higher. At the moment the FID output is 0.5 pA. Have lowered the purge flow to 20mL/min. The total flow on the inlet is where it should be. I could not find the jet restriction test on the FID diagnostics screen. I have lowered the H flow to the FID from 40 ml/min to 30 ml/min an back again.

Quite puzzled. Wondering when to call in for service.

Thanking you in advance for any and all additional help.

John Prohira

Evans Chemetics

Original Message processed by david 

Re:  - Re: Setting up splitless method on Agilent 7890B GC (09-Feb-2017 14:31)

Hi John,

Thank you for checking your inlet readings.  The total flow value is a quick 'sanity' check to make sure things are operating properly.   With that liner, the expansion volume of ethanol should be well contained. 

Keeping all other things equal, if you switch your inlet from splitless to split mode (~50:1 ratio), do you see a peak?  I would expect a huge signal with a 0.5uL injection of ethanol. 

Did the FID signal change when you adjusted the flows?  Were you able to see condensation on a surface?  While it isn't uncommon to see very low FID output signals on a clean system, below 0.5pA, I always question if the flame is truly lit.  It can run just fine at 0.5pA, but it is another one of those checks.

I apologize for my vague instruction on the jet restriction test.  Here are the steps:

To check for a plugged FID jet:

1Leave the column installed in the FID. If already removed, cap off the detector column fitting in the oven. (Leaving the column installed will determine if the column is installed too high in the jet, plugging the orifice.)

2Set the makeup flow to “Off”. Confirm a reading on the GC display of 0.0 mL/min for actual makeup flow. If not 0.0 – follow the procedure To Zero the EPC Module Pressure Sensor.

3Set the hydrogen flow to 75 mL/minute (increase the H2 supply pressure as needed to achieve this flow rate setting.)

4Monitor the makeup flow “Actual” reading

If the makeup flow is indicating a value in excess of 1.0 mL/min, this indicates that the jet is plugged or partially plugged; pressure is backing up from the H2 channel into the makeup channel of the EPC system, resulting in a false flow indication on the makeup channel.

Alternately, remove the jet from the housing and hold it up to a light source. Check the holes in the jet for contamination. If plugged, replace the jet.

Abbey

Thank you Abbey for your attention and suggestions.

At this point I am at a loss. Following are observations made this morning after reading your last input.

Switched inlet from split-less to split (50:1) and do not see ethanol's peak.

FID signal remains low, 0.1-0.2 pA. No condensation observed on a mirror. Perhaps the flame isn't truly lit.

Attempted a check on plugged FID jet. Did not remove it for visual inspection. Column remained installed. From GC keyboard leaving flow through column at 1.8 mL/min turned

off the makeup at the detector with reading of 2.3 mL/min without any H flow. Plugged inlet?

Perhaps its time to put in a service call. This is a new instrument, granted my prowess in the lab leans more towards LC, at this point I need an expert to evaluate the FID

and remedy any issue I may have caused during installation.

Thanking you in advance. Will place that service call on Tuesday but will await a response from you.

John Prohira

Evans Chemetics 

Original Message processed by david 

Re:  - Re: Setting up splitless method on Agilent 7890B GC (10-Feb-2017 18:21)

Hi John,

You are on the right track-- it sounds like your flame is not lit.   The reading you are seeing on the Hydrogen line is likely back-pressure from the column flow.  That result could indicate a blockage in the jet, but the flow is a little low to say for sure. 

It is always good to start by checking the available pressure in your gas tanks. 

You can start the re-lighting sequence by pressing the [Front Detector] or [Back Detector] keys from the GC front panel, then scroll down to Flame.  Press [OFF] then [ON].  If you scroll back up to the flows, you should see them change as part of the ignition sequence.  You should see your signal spike very high, then taper off to a value less than 20pA.  If this doesn't help, watch for error messages.   

The default FID settings are:

Temp:  350C (320C will work, too)

Air:  400mL/min

Hydrogen:  30mL/min

Makeup (He or N2): 25mL/min.

The flame should light under these conditions.

I find with new FIDs,gently blow across the top of the chimney helps them to light.  You don't want to be right over the chimney, as the flame will pop a bit.  I recommend keeping your face in front of the mainframe, then blow across the chimney(6" or so away).   It sounds silly, but it seems to work!   I would try to light the FID 3 times before moving to the next step. 

If the FID doesn't light, the next step is to remove the column, cap the FID base, and try to light again.  If it doesn't light with the base plugged off, and the flow readings are at setpoint, your next step is to replace the jet.  It it does light, it could be that the column was installed a little too high, which causes the ignition mix to be sub-optimal for feeding a flame.  Trim a few inches off the end, and re-install by gently inserting until the column bottoms out, then pull back a mm or two.  Tighten down the fittings and try again to light the FID. 

Abbey

Hi John,

You are on the right track-- it sounds like your flame is not lit.   The reading you are seeing on the Hydrogen line is likely back-pressure from the column flow.  That result could indicate a blockage in the jet, but the flow is a little low to say for sure. 

It is always good to start by checking the available pressure in your gas tanks. 

You can start the re-lighting sequence by pressing the [Front Detector] or [Back Detector] keys from the GC front panel, then scroll down to Flame.  Press [OFF] then [ON].  If you scroll back up to the flows, you should see them change as part of the ignition sequence.  You should see your signal spike very high, then taper off to a value less than 20pA.  If this doesn't help, watch for error messages.   

The default FID settings are:

Temp:  350C (320C will work, too)

Air:  400mL/min

Hydrogen:  30mL/min

Makeup (He or N2): 25mL/min.

The flame should light under these conditions.

I find with new FIDs,gently blow across the top of the chimney helps them to light.  You don't want to be right over the chimney, as the flame will pop a bit.  I recommend keeping your face in front of the mainframe, then blow across the chimney(6" or so away).   It sounds silly, but it seems to work!   I would try to light the FID 3 times before moving to the next step. 

If the FID doesn't light, the next step is to remove the column, cap the FID base, and try to light again.  If it doesn't light with the base plugged off, and the flow readings are at setpoint, your next step is to replace the jet.  It it does light, it could be that the column was installed a little too high, which causes the ignition mix to be sub-optimal for feeding a flame.  Trim a few inches off the end, and re-install by gently inserting until the column bottoms out, then pull back a mm or two.  Tighten down the fittings and try again to light the FID. 

Abbey