Chromatography Forum

I used new septa and column out of the shelf for the Leco Pegasus III with Agilent 6890 GC but got into trouble after about 20 injections. Split/Spitless injector Septa cracked and blocked the column and injector and I had to clean the injector. There are another new pair of septa and column. I was wondering what is the shelf life of the red septa and DB-5 and HP-5 columns shipped with the instrument. Mine was sitting in the shelf since 2002. Septas are expired I am sure since I saw what happened but I am not sure about the columns yet? What will happen if columns fail? Will I get contaminated MS chamber? Should I go and try them? I need to do a HCB sensitivity test.

I have taken columns off the shelf that have been several years old and they have worked fine. If the stationary phase has degraded, you will have poor retention characteristics and high bleed - just like and column that has become old. If the column coating has aged, the column will be brittle and will break as you flex it to install it in the instrument. (Not common, but I saw a "new" column come out of the box and break into many pieces. Bad polyimide was the only explanation that we could come up with.)

Actually there is no shelf life reported by the vendors because it strongly depends on the column storage conditions. If column has been stored in it's original box with well end-capped (originally end-closed or capped with a septa), it may work well, otherwise, oxygen and moisture entered into the column may oxidize the stationary phase and lead to high bleed and so MS contamination. So Be careful about what column you install.
Column bleed also increase noise and leads to low sensitivity. Column stationary phase (HP-5 , DB-5 , ...) has no effect on the sensitivity test, if there is no bleed.

Septa exposed to the air may be cracked easier than those stored well.

Is there anyway of testing column bleed?

I tried to do this by pulling out the injector end of the column and I did leak check but didn't see any leak in the system. Could this be because column is blocked (due to aging or debris of septa went into the column)?

Can't you assess any column bleed by doing a noise evaluation?

I haven't myself noticed any column degradation by aging if stored correctly.

I have come across septa that were bad. About 30 years ago, we purchased some gray septa from HP, and these would fail to seal after a half-dozen injections. We complained to HP, who then did some testing and found out that the batch of resin from their supplier was off. We were just the first to notice and supply information back to them.

My problem is I don't see HCB 2ng peak in TIC. Even can't see 20ng peak. 200 ng peak hardly seen in TIC. I feel something is wrong or I am missing something here. Manual says with 2ng I should see 10:1 signal/noise but I see almost no peak. I changed septa and column but seems didn't help. Any other advice to pin point the problem?

Column bleed is the normal elution of stationary phase as the GC column is temperature programed. All columns will show a certain amount of bleed as the oven temperature increases.
The question is whether the bleed is normal or excessive for your column and conditions. Column bleed is characterized by a gradual baseline rise which reaches a plateau at the final temnerature of the program. This rise typically begins approximately 20 to 30°C below the maximum operating temperature of the column. The stationary phases used in capillary columns are susceptible to oxidation. If the column exhibits a high baseline rise, the column may have been subjected to oxygen or high temperature.
To determine how much column bleed is acceptable for an analysis, you must have an understanding of the necessary detection limits, the type of column being used, and the detector and signal sensitivity during operation.
After column conditioning, run a blank and investigate the baseline rise.

For leak test, you can consider m/z= 18 (water),28 and 32 (air). Acceptable levels of air and water vary from system to system, but usually If there is high leak, intensity of m/z=28 goes higher than that of m/z=18 (m/z 32 also increases) . By removing the column from injector the intensity ratio of 28/18 increases severly after a few minutes (notice this is not used for leak test). In your case, the column may be blocked by the worn out septa. Cut 10-30 cm of column and install again. Clean liner befor column installation.

Don't expect to see the 2 pg HCB peak in the TIC; there is too much noise. Select m/z 284 and 286 to display. (I take the two ions because the indiviual ions show up in various silanes and such, but rise and fall together in HCB. The instrument specification for signal to noise is for the 2 pg HCB peak - on the 284 trace.

Finally I got 10 pg HCB on 284 trace using DP method (SN:10 and detector voltage 1800). Any advice what else to do to get 2 pg HCB ? I used all parameters as manual says.

Also instrument is too noisy. Anyone has attempted to decrease the noise some how? I saw in new Pegasus models there is a large pad like something in front of the front panel opening to stop the noise I guess.

There is a sound baffle that can be placed on the front of the Pegasus to reduce noise from the fans. It can be purchased from LECO.

If you get a s/n of 10:1 for 10 pg HCB under a set of conditions, 2 pg HCB would be expected to give s/n of 2 under those same conditions -- lost in the noise. At what voltage does the instrument tune? If the instrument tunes below 1600 v, I would examine the GC and column installation into the transfer line. Be sure you are using the type column used on the original installatin and use the method parameters (GC, and MS. DP may have changed slightly with if you have anewer version of ChromaTOF ) as used in the installation check-out. If the instrument tunes above 1600 v, you may have an issue in the mass spec - and with what I understand the age of the instrument to be, the detector is one very real possibility.

Instrument tunes at -1350 V. Detector is almost new. That is why I am surprised why I am not geting 2 pg HCB. Other than HCB which is too old I couldn't think of anything else. I used the aged DB-5 column and I am not sure about its storage condition. About Chromatof version it is version 2.0 (pegasus driver version 2.11 .

With that tune on the MS, the MS is in good condition. Look at column installation, inlet conditions, and method parameters. The column should extend 3 to 5 mm beyond the end of the transfer line. I don't remember the spec for extention into the inlet. Peak width for the PCB peak should be about 1 or 2 seconds (check the specifications for the expected peak width - if you have a wider peak width, you have a chromatographic issue) and not tailing - best observed on a higher level injection. A new linlet liner with wool or open works. be sure you are using a new o-ring to seal the liner in the inlet. (Or if you are using a graphote o-ring, be sure that it is snug on the liner.)

Depending on your version of ChromaTOF you select peak width and number of points for averaging in the data procssing method or "automatic." Peak width is measured at baseline. If your version of ChromaTOF allow you to selct automatic for selecing points to average - do so.

If you are not sure that you have method parameters that were used wihen the instrument was installed and checked out, call your LECO service office and see if they will send you parameters. Be sure to tell them which version of ChromaTOF you are using. GC and Autosampler parameters should not change. Mass spec parameters will carry forward OK from across revistions of ChromaTOF as will most data processing parameters. But, I don't remember if there is anything that gets confusing if you use parameters from the current version of ChromaTOF and try to set up on an older version. (These days I'm looking at ChromaTOF, Mass Hunter, Xcalibur and probably somethign else -- and there you do find parameters that do not exactly match.)