How to eliminate ghost peaks?

[max]

Hello to you all! Is there anybody out there giving me some suggestion? I write from Italy, where I work as research assistant at ENEA research institute.
I'm just approaching to the GC world, I'm currently doing GC determination of N2O (RT=1.7 min) with a Varian GC 3300; I use a 2 mt.Porapack Q (80-100 mesh) packed column, kept at a 40°C temp., while the injector is set at 100°C. The instrument is equipped with an ECD, the carrier gas is N2, the flux is set to a value of 40 ml/min.
Since some weeks, I observe two ghost peaks with RT = 0.40 and 14.00 min, respectively. The first one is very tight and defined, whereas the other one is broader. I replaced the column and the septum but the problem has remained;furthermore, I noted that the two areas are proportional to the injected volume. That led me to suppose that the problem is due to an injector contamination, so that's the question:

1) Am I right?

And if so:

1) How to clean the injector (An "on-column" one)?
2) I read on internet that it is possible to eliminate impurities fro the injector simply disconnecting it from the column and increasing the temperature up to 350°C for several minutes; does ir risk to damage the injector?

Thank you all for suggestions

Max

[Rafael Chust]

Dear Mark:

Everything points to injector contamination.

There are available (and very cheap) injector cleaning kits in the market.

Basically, you will scrub the injector (column off, please!).

The idea of high temp (baking) the injector is valid and, as far your system allows that high temp, no damage will be caused to the system. Do not forget to keep the carrier gas flowing and the column off!

Hope this helps!

[max]

Obrigado, Rafael!
Injector contamination is the first explanation I thought about, but as I'm new with the instrument, I preferred launching a "message in a bottle" before moving.

Thanks a lot

Max

[chromatographer1]

By what means are you injecting the sample?

You may be seeing a pressure pulse for the 0.40min peak or a balance gas peak.

Have you run a blank injection to see if the ghosts are still there.

There may be an impurity in your sample for the later peak.

The big question is:

are the ghost peaks there when you run a blank?

If they are real, you have no hardware problem, but a sample problem.

[max]

Hello, chromatographer1,
The peaks are present even running a blank injection, and their areas are proportional to the injected volume. I also tried to start a run just putting the needle of the syringe in the injector, with the plunger full down, and the result has been:
1) No 0.40 min peak (so that could confirm your hypothesis about a pressure pulse)
2) A smaller 14 min broad peak.

What do you think about?

Thanks, anyway

Max

[Rafael Chust]

So injector problem will be...

[chromatographer1]

I leaned from your reply:

1. You are using a syringe and not a valve for injection.

2. You do not see the 0.40min peak if no gas sample from the syringe is injected.

3. The late peak gets smaller.

What to do?

Take your syringe apart and place in a heated oven for several minutes.

Insert the needle of the reassembled syringe into another heated injection port and leave it there for several minutes.

Then leaving the plunger down as before, perform an 'injection' .

Repeat, but this time leave the syringe in the injection port during the run.

Repeat injecting dry nitrogen or helium (whatever your sample has for a balance component of the mix you are using.)

Repeat injecting your sample again. Could the late peak be water vapor?

Please report your findings and then let's discuss.

[max]

Ok Rod (Is that your name?), I will follow your useful suggestions. See you tomorrow (Here it's 4:30 pm) for discussion about the results.

Thanks

Max

[max]

OK, I didn't follow the protocol suggested by Rod, but I have increased the column temp. up to 100°C, I have injected air, and that's what I got:

a) 0.40 min peak has remained the same, maybe it decreases towards the baseline more slightly, but that's all;
b) RT of the second peak has decreased down to 1.80 min.

My conclusion is that the first peak is Oxygen, and the second one is water vapour, according to Rod's latest hypothesis. Really, I joined the group just 2 months ago, after the trouble appeared. I just received the work to "resolve" it! Even if it seems to me strange that, earlier, non water vapour peaks appeared on the chromatograms, especially thinking that the sample is kept by serum bottles anaerobically inoculated with lagoon sediments and water, in order to assess denitrification ratio. Maybe, using different column, the peak was so broad to not to disturb the chromatogram. I think that there are 3 solutions

a)Room temp. injection, so that water gets out quite broader. I don't like that;
b)Setting up a program with an increase of column temp after N2O gets out; not bad, but it takes to wait for the system to reach an equilibrium after coolin'
c) Finding out a column temp between 40° and 100, so that water peak gets out earlier than 14 min and the N2O one is easiliy detected.

Hello to you all, and thanks for suggestions, they've been quite stimulating!

Max

[chromatographer1]

I am glad you believe you have a solution.

Sometimes a 'jump to a conclusion' solution has a happy ending.

Your early peak is probably O2 or CO (Any NO present should be about 0.6-0.7 min) or a pressure flow pulse disrupting the baseline caused by the syringe injection. CO2 should be eluting just before the N2O.

But it is good to systematically go through a trouble shooting process, both to understand what is going on, and to gather additional information about the problem.

best wishes,

Rodney George
Senior Research and Development Scientist
Gas Separations Research
Supelco
595 North Harrison Road
Bellefonte, PA 16823

[max]

The result of the earlier tests with different column temp, is that N2O amount in my samples is so little that it takes to inject large volumes 200 ul or more. Consequently, O2 peak is really large when using temp >40°C, and it covers N2O one. So, I'm thinking to se te column temp to 100°C after N2O gets out.
The real problem, in my opinion, is that the column equipping my GC doesn't fit with my aim; too short (2 mt), and the materiali (Porapack Q) strongly keeps water. Is there anybody out there who can suggest me a different column type wich is good for N2O, I mean, longer or made with a different material?

Thanks

Max

[chromatographer1]

You can use a Porapak P which will not retain water as long.

Or a Supel-Q PLOT capillary column, 30m 0m53mm ID from Supelco.

(my company)

Rod

[max]

Thanks Rod, I will soon search for informations about a Porapack P packed column. If your company sales columns like that, what are the prices?

Max

[chromatographer1]

In order to minimize the width of the Oxygen peak, you may wish to increase the ID of the column in the analysis. The "P" polymer has the same separation as the "Q" polymer beads but is faster with less separation. A longer column is also advised, as well as deactivating the tube used if you still wish to use metal columns instead of glass.

Beware of ECD sensitivity changes with a large injection of oxygen.

Good luck and best wishes,

Rod

Supelco
595 North Harrison Road
Bellefonte, PA 16823

http://www.sigmaaldrich.com/Brands/Supelco_Home.html

sales 800-247-6628

tech service 800-359-3041