Chromatography Forum

Dear Chromatographers.

Recently we ran into linearity problems when analyze VOCs in soil.
Our system is composed of:

* A GC/MS - DB624 column

<< split 4:1,
<<4.8 ml/min He,
<<35°C init. temp.
<< 1.2kV
<< total flow 24 ml/min

* Atomx purge and trap unit, running in soil mode with the default soil method:

<< Valve temp = 140°C
<< Transfer line = 140°C
<< Sample vial = 40°C
<< Soil valve = 100°C
<< Standby flow = 10ml/min
<<water added = 10mL
<< stirring bar speed = medium.
<<purge time = 11 min
<<purge flow = 40ml/min
<<trap temp = 20°C ( 250°C when desorbing)
<<condensate trap = 20°C
<<dry purge time = 2min
<<dry purge flow = 100 ml/min
<<dry purge temp = 20°C.

* Consider this:

<< We inject 2uL of methanolic standards directly into the GC injection port, 1,2,4,16 & 32 mg/L, linearity....OK.

<<We purge water standards ( Atomx in water mode, 5mL fritted sparger) of 0.2; 1; 2; 5 & 10 ug/L, linearity....OK.

<<Our soil technique is: Add 5mL of water to a 40mL VOA amber vial, then we add a stirring bar.
After that, we inject quickly (under the water level) one vial at a time: 50, 80, 100, 150 & 200 ng from a methanolic standard. Linearity...too bad.

For instance:

Toluene
50ng = 43K area counts
80ng = 47K
100ng = 193K
150ng = 256K
200ng = 189K

Any help would be appreciated.
Best regards.

How long do you allow the spiked sample to equilibrate before you purge it, and do you stir or agitate it during equilibration ?.

Peter

I add my surrogates and internal standards (dissolved in MeOH) to 5mL of 18.2 MegOhm DI water in a gas tight syringe. I add spikes if needed, at the same time, all in the same 5mL syringe. Then I open the 40mL voa vial containing the soil and empty the syringe into it by letting it run down the side. Quickly but without turbulence. The Archon adds an additional 5mL of DIW just before it gets purged. Methylene Chloride recovers very well in this fashion. Cal curves are put on in the same way. RSD's are typically below 10%.

I should add that when we first got an Archon with Tekmar 3000 we let the Archon add the internal standards and surrogates. The results were terrible, very non linear. Possibly it was a valve problem. But we started doing it by hand for each sample prep and its been fine for linearity ever since. We run both waters and soils in soil mode.

My thought would be that by having surrogates and spikes all mixed in the gas tight syringe; perhaps any volatile losses are compensated by all having the same exposure. That said, RSD's for internal standards are quite tight. So, that would argue that losses are very minimal.

4:1 split? What are your column dimensions?

* A GC/MS - DB624 column

<< split 4:1,
<<4.8 ml/min He,
<<35°C init. temp.
<< 1.2kV
<< total flow 24 ml/min

Is that 4.8 ml/min through the column into the mass spec? If so you are near the limits of what the best vacuum systems can hope to evacuate, except for maybe a triple quad system.

I normally run about 1ml/min through either 0.25mm or 0.18mm columns and use at least a 40:1 split ratio so I have a good flow from the purge and trap during desorb. Overloading the MS or too fast a flow in the column could be the problem.

James_Ball wrote:
* A GC/MS - DB624 column

<< split 4:1,
<<4.8 ml/min He,
<<35°C init. temp.
<< 1.2kV
<< total flow 24 ml/min

Is that 4.8 ml/min through the column into the mass spec? If so you are near the limits of what the best vacuum systems can hope to evacuate, except for maybe a triple quad system.

I normally run about 1ml/min through either 0.25mm or 0.18mm columns and use at least a 40:1 split ratio so I have a good flow from the purge and trap during desorb. Overloading the MS or too fast a flow in the column could be the problem.

I'm with James here. 40C initial temp 0.18mm DB-624 or DB-VRX column, 1mL/min flow, 40:1 split ratio. Ramp to 1.3 mL/min near second IS. Then ramp up to 1.5mL near the end of the run to "run off" heavies but back down to 1mL/min for the next injection. Whole 8260 run takes 18 min.

[/quote]I'm with James here. 40C initial temp 0.18mm DB-624 or DB-VRX column, 1mL/min flow, 40:1 split ratio. Ramp to 1.3 mL/min near second IS. Then ramp up to 1.5mL near the end of the run to "run off" heavies but back down to 1mL/min for the next injection. Whole 8260 run takes 18 min.[/quote]

I normally run the same thing but with constant flow of 1ml/min all the way through 35c hold 1.5 then 9c/min to 160c then 25c/min to 200, total run 17 minutes and that gets the trichlorobenzenes off with time to spare. Split can be 60:1 and have detection limits as low as 0.5ppb for most compounds.

Dear all.
Thanks for your replies.

*The column is a 75m x 0.53mm x 3um. For the moment we cannot use a narrower column. Nevertheless, curves in water mode are OK.
*We dont equilibrate the vial before purging because the stirring bar, flow of gas and the 40°C purge temperature, may be enough, I think.

Right now we are analyzing a new calibration curve. Yesterday the three stage needle was thoroughly cleaned.


Best regards.

Are you splitting after the column in addition to the inlet split? Otherwise that is too much flow for most mass specs. Thus, most of us now use .25 or .18 columns with a much larger split ratio. Newer instruments don't handle the water as well as the older ones.

It looks like you are purging 50 ng for your water and 200 ng for your soil.

You need to equilibrate after spiking, that's why I asked about it four days ago.

Peter

Peter Apps wrote:
You need to equilibrate after spiking, that's why I asked about it four days ago.

Peter

The longer an internal standard sits on soil the lower the recovery, unless it is sand. Most autosamplers add the internal with 5ml of water then preheat to 40c for about 0.5-1 minute then start the purge. Usually get good recoveries that way.

OK, Im going to equilibrate after spiking 1min @ 40°C.

Best regards.

...just out of curiosity,

does a 0.25mm DB-624 or VRX column resolve Vinyl chloride from solvent? without the use of a cryogenic or peltier element.

Best regards.

James_Ball wrote:

Peter Apps wrote:
You need to equilibrate after spiking, that's why I asked about it four days ago.

Peter

The longer an internal standard sits on soil the lower the recovery, unless it is sand. Most autosamplers add the internal with 5ml of water then preheat to 40c for about 0.5-1 minute then start the purge. Usually get good recoveries that way.

Which confirms equilibration as the likely cause of the problem; grabbing a snapshot works OK if everything is automated and the IS is the same amount every time, but if the addition is manual (and offline if I understand correctly) and the standard amount varies, as it must for linearity, then both repeatability and linearity are going to suffer.

Peter