Reproducibility Issues after system maintenance

I am using a splitless method to analyze a solution of a material that contains non-volatile polymer and volatile monomer I am running a sequence in which use one method to  volatalize and analyze the monomers and a few  high boiling solvents followed by a second method that degrades the polymer and removes it. I can see that method 2 is pretty effective in cleaning out the liner from the polymer and resulting contamination. We tuned the method untile we had very nice overlaying chromatograms and a very good calibration result.

Then I replaced the filament, the guard column and the MS transferline (both fused silica, deactivated) and all the the connections and ferrules from these to the MS, the PUU (which is used to connect to the analytical column on both sides) and the inlet. After that the calibrations still look fine (area ratios of standard to volatile components), but the overlays don't look good anymore. I even observe shifts in retention time.

When I use split methods for other materials, the overlays look much better.

There is no leak based on the tune and the tune looks very good in general. 

I replaced the liner+O-ring, septum and syringe step by step. Still the same issue.

I also just recently replaced the vent line filters.

A consultant of ours suspects the vent valve to maybe not seal properly. Also, I was wondering if it could be the gold seal that needs to be replaced. 

Also it would be great if someone could explain to me a bit more how the sample is injected and volatalized and then released from the inlet in a split/splitless mode. How long is the material heated and not released in the inlet, How long does the vent valve open before it closes again? How far does the syringe go into the liner? Is there a part of the liner that is not heated as much or is the whole liner heated homogenously in the inlet?

My system is an Agilent GC 6890N - MS 5975C. 

I used both of these liners: https://www.agilent.com/store/productDetail.jsp?catalogId=210-4004-5, https://www.agilent.com/store/productDetail.jsp?catalogId=210-4022-5 I know they say split, but I was told they both work for splitless as well and since I have some split and some splitless methods that worked pretty well for us.

Thanks. Bettina

Parents
  • Please read up on split and splitless along with inlet maintenance in the GC manuals. 

    There are many resources on Agilent.com and here on the Agilent Customer Community.

    Maintaining Your Agilent Split/Splitless GC Inlet

    Maintaining Your Agilent Split/Splitless GC Inlet - same title, different presentation!

    Understanding the GC Inlet. Which one is more appropriate for your method? (agilent.com)

    (+) Column Installation - Split/Splitless and Multimode inlets =and= SQ and TQ MS transferlines - Files - GC/MS - Agilent Community

    There are some easy questions to start with.  What is your column configuration?  Is it set properly in the software?  You have a guard column, then two columns with a PUU and going to vaccum in the Mass Spec, is the software and method configured properly?  Does column 2 have about 0.1 to 0.2 ml/min MORE than column 1?  This is necessary so that there is always positive flow from the PUU supply so that sample can't go up the PUU gas supply line.    What is your splitless  Purge ON time? What is your inlet temperature and initial oven temperature and initial oven hold time?

    The cross gold seal is not necessary.  The cross seal is only required if you are running in modes with very high inlet flows, like split over some few hundred ml/min, or have samples with so much residue that collects on the gold seal that it can become clogged  (this would mean that you should do dramatically more inlet maintenance). 

  • Thank you very much. Will work through this. I am not totally sure about the column installation. Our system is set up in a way that the flow that goes though the sequence of guard column-PUU-analytical column-PUU-MS Transfer Lines is viewed as one column as far as I understand, so there is only one column and I cannot set the gas flow differently for the different parts of it as it goes from the inlet into the guard column which is connected to analytical column (through 1st PUU), which is then connected to the MS transfer line (through 2nd PUU) and then to the MS. All 3 columns have the same dmensions. Guard column and MS transfer line ar deactivated fused silica. Maybe I need to reconfigure that. I'll look into that using the resources you shared.

    And yes, we are injecting a polymer that is non volatile and gets stuck on the liner. It is possible that some of that is collecting in the liner as even burn-outs of inlet and column do not fully get rid of residual peaks.

    The column has been used for a loooong time and this material has been injected many times over these past few years. We had a PM done by Agilent last year and I'm pretty sure they cleaned the inlet, but since then we did lots of development work which involved the injection of non-volatiles and compounds with very high boiling points of which I also see impurities in the blanks.

    I may have more questions once I went through all the material you shared

  • I think I understand what you mean and I used the wrong term. We actually have this Union: https://www.agilent.com/store/productDetail.jsp?catalogId=G3182-60581&catId=SubCat2ECS_896718. It is not a PUU, but a UU. Correct? Sorry, I am not an expert (as you can see :-))

    I'll sent all the info you asked for tomorrow.

  • A PUU is a Purged Ultimate Union and has flow. The UU is an Ultimate Union - a connector.  If you have three column pieces joined by UU, just make sure that the software knows how long the total is.   

    The easiest way is to calibrate the column.  It can be done in manual tune by watching, or by setting up a run and using the resultant chromatogram.  

    In Manual Tune, set up repeat profile of 28, 28, 28 -- nitrogen.  With the filament on and the PFTBA off, inject 3 uL of air and measure the time it takes for the nitrogen peak to come out to the nearest second.  Write that number down.

    In a run, setup the method with 0.0 minute solvent delay, scan 10-100, inject 5uL 20:1 split, and put an empty vial into the autosampler.  The resultant chromatogram should have a peak. Integrate it and write down the retention time.

    In the GC Edit Parameters page, Configuration, Columns, Choose the column installed so that it is highlighted in blue and then click on "Calibrate".

    This window comes up:

    Then you choose "If unretained peak holdup time is known --- and click on "Calc Length..."

    Enter the Holdup Time of an Unretained Peak, click elsewhere in the window or hit tab and you'll see the calculated length.  Select OK, OK, and then the updated calibration should be entered.  

    The benefit is that if you cut off a bit, you can do this test again and it will set the proper pressures for the required flow for the new length.  Calibrating the columns makes it easier to keep your retention times the same over time with column changed and column maintenance.

  • Hi Paul, How do I set up the repeat profile of 28, 28, 28? Do I enter it as mass 1, 2, 3?

    What method do I load when I inject 3 uL of air? I would load an empty vial and select 3 uL Injection volume, but what should the method be? 

  • I assume I can inject 3 uL of air manually, but I don't really know how to do that. I just know how to do that with the autosampler

  • I tried that and there is a peak at 28 already. It did not increase when I injected 3 uL with the syringe by hand into the inlet

  • You would remove the ALS and manually inject through the septum with a handheld syringe.  However, one way to use the autosampler to do this is to squirt air duster into a vial and cap it off quickly. The sampler will pull up enough air duster gas for you to be able to see a peak.  Set your MS method to scan from, oh, 15-125 or something like that and you should see a peak in the chromatogram that you can use for unretained volume calculations and column calibrations.

Reply
  • You would remove the ALS and manually inject through the septum with a handheld syringe.  However, one way to use the autosampler to do this is to squirt air duster into a vial and cap it off quickly. The sampler will pull up enough air duster gas for you to be able to see a peak.  Set your MS method to scan from, oh, 15-125 or something like that and you should see a peak in the chromatogram that you can use for unretained volume calculations and column calibrations.

Children
  • Thank you Paul. I was able to do all that now. I also did inlet maintenance:

    opened the inlet completely, replaced guard column+fittings (ferrules on both ends, column nut to Ultimate union), gold seal+washer, and inlet reducing nut (used a better looking one from back-inlet), septum retainer nut (used a better looking one from back inlet), opened purge lines which looked clean, cleaned inlet with solvent, replaced vent line traps. Put everything back together and performed inlet leak check which is leak-free. Fixed a leak at the Union/Guard column connection, checked for leaks at inlet/guard column connection. seems leak-free

    If I test termperature: 70C, Mode splitless, purge flow: 50 ml/min, column flow: 4 mL/min, and click prep run twice the total flow fluctuates around 6 mL/min. Tune fails - does not complete. I looked through everything and cannot find an issue. 

    I gues one thing left to do is replace the purge lines. I have the Cu tubes here (G1544-20610), but would like instructions if possible. I could not find this in the manual.

    Is it possible that the valve is malfunctioning?

    Best, Bettina

  • Bettina, turn the column flow down to 1.2 ml/min and try to tune.  4ml/min is much higher than the optimal flow for the ion source.

  • Thank you so much Paul. I did not understand that this is important. Generally, Will I need to tune the instrument for each method settings and use the associated tune to get meaningful data or can I do a general tune that I can use for all methods? 

    The tune completed now, still showing N2/O2 at 14.2/5.5%. So, either I need to give it more time after fiddleing with everything or there is still a leak. 

  • The total flow is around 6ml/min.... do you have the Septum Purge Flow Mode setup for Standard or Switched?

    If it is in Standard mode, the Total Flow during the splitless time is equal to the column flow plus the septum purge flow.

    If it is in Switched mode, the Total Flow during the splitless time is equal to the column flow.

    Please remember that column flow is always calculated from pressure/temperature/restriction of the column.

    The copper split vent line is easy to check when you have the inlet apart for cleaning. When the column and liner are out and the gold seal removed, you can squirt solvent from the disconnected split vent trap end and it will drip out of the bottom of the inlet into the oven.  Since it's all back together clean and leak tight, it's more work to do it again.  Replacing the copper tube requires removing the autosampler bracket to get access. Other than that it is typical 1/8" swagelok fittings on the inlet body and the split vent trap body.

    Do you have a flowmeter? Measure the flow out of the split vent and compare it with the setting on the GC in Split mode.  If it tracks, the valve is probably fine. Sure the valve could be malfunctioning, but it is not a high failure part.

    What liner are you using in the inlet?  There are liners out there that are just a bit big outside diameter. They can restrict the flow and cause this, too. It's certainly worth putting a split liner inside to test just to be careful.

  • Set all methods to use 1.2 ml/min of column flow into the MS and then tune at that flow rate.   In general changing column flow is not the best way to run a GC column, especially into a Mass Spectrometer.  Set them all to 1.2 ml/min and then change your oven temperature ramps and timing to alter your peak's retention times.   

    If you just have to change column flows, then yes, the tuning changes depending on the number of helium molecules coming out  - the ionization chamber's instantaneous pressure/vacuum, and so it is best to tune at the same exact flows and temperature conditions you'll be using in each method.   But tuning compound is just like samples and leaves residue on your source as well so the more you tune, the more you have to clean it.  It's a lovely dance.  Try to tune only when the system needs to be tuned.  Most operators do not need the system to be at the most perfect tune to achieve good data anyway.

    N2:02 in air is ~4:1. That's not the ratio you're seeing, though. how long has it been pumping?  Don't try to do anything to a GCMS for at least two to four hours after pumping down from being open.  Make sure you have no leaks before the GC, too.

  • Thanks. :-)

    I never cleaned the source. It was last cleaned during a vendor PM on 01/27/2021. Our SOP (set up by previous scientist/tool owners) is to run the tune once every day to check for air leaks and make sure that N2/O2 is below 5%. How can I check for that without using a tune? And yes, I noticed that I am not seeing this ratio I would expect for an air leak. Our Helium cylinder is replaced when the main pressure is down at 600 psi. Currently it is at 700 psi and we are waiting for Helium. The outlet pressure of the cylinder has been constant though which is why I don't think that that is the problem. What else could this be. Is it possible I need to clean the ion source? Could I bake it out? I cleaned the inlet with acetone. Could this be an issue?

  • It's a dance with many partners!  Would you please share the latest tune report?  You said "pumping for 2 days" below. What is the vacuum gauge reading?

    The N2:O2 ratio could be from some combination coming from your gas, leak or leaks, screwed up tune parameters due to too high flow, too low flow, poor vacuum, and on and on.  

    After a GCMS is tuned, the simplest way to test it is to have a known peak of interest that you inject. Then you evaluate the chromatogram for retention time, chromatographic peak shape, and height/area - which checks mostly the sample, syringe, inlet, flow rates, column, heaters, etc, but also the MS sensitivity.  You check the spectrum of the peak for relative abundances of ions across the range.  If you set the range to start around 15amu, you'll also see water, nitrogen, oxygen, etc.   So one injection takes a good picture of the system right then.  If that peak of interest is in the right place, at the right time, is in the typical range of height/area plus or minus whatever you deem acceptable, and the spectrum is correct plus or minus whatever you deem acceptable, the system is ready to run.   This is how some EPA methods work. You inject DFTPP, evaluate the peak and if it meets criteria the system is ready to go.  That tests more than just the ion source.

    In that test if the spectrum is starting to tilt, losing higher masses relative to lower ones, or the total sensitivity is lower than your limit, then run tune.   Some operators running clean drinking water applications only tune once every six months or more!  

    I doubt you need to do anything with the MS right now, but when you're ready to clean the source you can see videos on how to do it here:

    Eliminate the Fear of Mass Spec

Was this helpful?