Recurring ghost peak

[Chris Jones]

Hi everyone,

Been a long while since I consulted the brains trust that is ChromForum. Mostly been busy with other stuff, but I have recently fired up the Shimadzu GC2010 with FID and I've been trying to get it into a fairly normal baseline. New DB-WAX column (30 m) installed, new splitless liner, new septa. I am just conditioning the column so I inject an air sample, 40 degree oven for 10 mins, then up to 200 for 10 min then 230 for 20 mins.

I am constantly getting a peak at about 35 minutes. I have seen this with different columns, different gas bottles, different solvents. Injecting air gives the same peak. I have concluded that the septa might be coated in a waxy compound, of say C15-C20. New septa makes no difference, so I'm confused. Any ideas?

Cheers guys,

CHRIS

[Peter Apps]

Hi Chris

You might be getting a contaminant from the septum, so you need to do some troubleshooting to eliminate other possibilities.

First: how much air are you injecting, and why ? Air is invisible to the FID and so you cannot be measuring the holdup time, and there is a faint possibility (depending on how much you inject and the inlet temperature) that you are degrading a little bit of inlet deactivation or column phase each time, and that the peak is the product.

Do you get the peak if you run the temperature programme without doing an injection at all ?

Do you get the peak if you do an injection, but with a volume of zero - in other words the syringe penetrates the septum but does not inject anything ?

Is the peak bigger after the GC has been in standby for a while, compared to in a run that quickly follows the previous one ?

Do you have scrubbers on your carrier gas line ?

Check the septum purge flow.

Peter

[Chris Jones]

Hi Peter, thanks for the reply.

I am injecting air to comfirm that the solvent doesn't contain a contaminant. I'm only injecting 0.5 microlitres at a time. I am doing some more tests to see if the problem is injection related, including no penetration, and needle penetration only. If it turns out that needle penetration results in the peak but not anything else, then I will start to play around with the injection volume to see if it's my syringe that's causing the grief.

Also, the baseline rises with temperature pretty substantially. I try to bake the FID at 400'C and it gives a characteristic 150,000 uV rise and fall, but there is still a good 500 to 1000 microvolt rise with the increase in oven temperature (now 40 to 230 at 10'C per min, held for 20 min).

I'll report back later

Cheers,
CHRIS

[Peter Apps]

Hi Chris

1000 uv of drift is pretty normal for a wax column under temperature programming.

Peter

[Mr.Buffer]

Might your air sample contaminated? I had same problem in the past while i was working GC/HS method. Then I purged vial with nitrogen and tried to inject it again. All ghost peaks were gone forever.
İf you are using GC injector method , you might consider to change your injector with new one cuz it might be contaminated.

Best regards.

[Don_Hilton]

Let us know the results of the tests Peter listed.

By what means have you confirmed that you have flow through the column - and at the correct flow rate? (an injection of butane methane from natural gas can be good for this - or in a pinch, butane from a cigarette lighter and a hot GC oven. Reading a pressure gague shows that there is pressure in the inlet which should reslult in the correct flow.)

[Chris Jones]

Hi All,

I have been injecting 0.1 ul of air from a sealed vial with nothing in it (never had anything in it). I am not rinsing with any solvents either. The injection port has a new septa and the liner is brand new. The column is brand new too.

I have even changed the syringe and I still get the same contaminant (verymuch a grease like compound, based on it's retention).

I'll summarise the results in a bit, but it's looking like the septa is causing the problem. Maybe I need to find a different brand of septa?

[Chris Jones]

Still there.

For the record - this is a brand new column, brand new liner, brand new septa, brand new needle, never been rinsed with solvents. It is a 0.1 ul injection of air (it won't do 0 ul).

Next step is to do a series of repeats to confirm it is either decreasing or increasing... Yes, it does seem to increase in size the longer the GC sits idle

[Don_Hilton]

Yes, it does seem to increase in size the longer the GC sits idle

Indication that something is collecting on the head of the column over time. What grade of carrier gas are you using and what types of inline filters or scrubbers are you using? The plumbing from the gas bottle to the instrument should be a regulator with a metal (not rubber) diaphragm, cleaned tubing (suitable for GC) connected with compression type fittings (not soldered). If you have an old regulator of unknown history, refrigerator grade tubing, or soldered joints, expect to have stuff migrate into the GC and cause trouble. And even if you are using all the right parts and assembly techniques - a trap close to the back of the GC is always a good idea. I use the fancy indicating traps. With research grade gas, the traps stay in place "forever" until I open that one rare bad cylinder of gas. Bad cylinder detected - column and GC internal plumbing saved.

[Chris Jones]

It claims to be UHP helium, and not cheap either. But I will check to see that the regulator has a metal diaphram.
I have some more data coming in now.

[Chris Jones]

OK, yep, it looks like some crud is accumulating on the glass wool of the liner or possibly the septa, because when the needle is inserted it seems to dislodge it. The next injection was identical to the last, but the crud peak at 24 min is now much much smaller.

New scrubbers and new gas supply, better regulators???

[Chris Jones]

AHA!

This next run was a program without any injection whatsoever. And we still get the damn peak! The trace in Pink was the no injection run, and the peak in black was the previous 0.1 ul air injection.


So basically, we have a problem with the carrier gas depositing crud on the liner, but somehow, even without injection, it manages to move to the column and elute at the same time every time.

Stranger and strangerer...

[Peter Apps]

Hi Chris

The no injection run clearly absolves the syringe.

There are three possible sources for crud that accumulates at the top of the column; carrier gas, septum bleed, inlet packings.

Carrier gas has already been discussed.

Is there any way that you can replace the septum with a metal seal (I can do this on my Varian 1079 inlet, I odn't know about your instrument) ? If not then cool the inlet completely, and do two runs.

Take the glass wool out of the liner and do two runs.

Peter

[R13]

If You make several "no injection" runs one after another - does contaminant peak intensity change (decreases, remains the same)?

Have You checked septum purge flow (if the flow from vent is in accordance with the setting)? Did You try to increase septum purge flow?

Change to another brand/type of the septum might help to identify the source.

[Chris Jones]

Hi again everyone,

I have replaced the regulator with a brand new one and have also changed the He scrubber. The temperature program is 40'C for 1 min, then 10'C per min to 230 'C held for 20 mins. Pressure is constant at 80 kPa and the column flow is a constant 1 ml / min. I have been doing some testing without making any injections and these are the following results:


After these I decided to do another run.


So basically, the baseline noise comes down with time (good) but the ghost peak continues to show up, and it seems to get larger with the machine being idle (He flow is low, occasionally off).

OK so what are we left with? Not the traps or the regulator, not the liner or septa. If no injection results in a peak (of greasy properties, similar to that of say a C18 aliphatic hydrocarbon) where the hell is it coming from and how is it accumulating?

Oh, I doubled the purge flow from 3 ml/min to 6 ml/min for the last 3 runs and I can't see any real difference.

Any ideas?

CHRIS