GC/MS Large volume injection

[James_Ball]

Has anyone tried the LVI procedure listed in this tech note?

http://www.restek.com/Technical-Resourc ... /ev_an1331

The LVI procedure is listed in the GC/MS Conditions section a few paragraphs down. They are using a normal split/splitless injection port with a 0.53 retention gap and cold oven technique with a 12.5ul injection.

We are looking at using the Agilent Multi Mode inlet to go to LVI for our semivolatile analysis but I am wondering if this would be a good idea to use in the interim to begin the setup process and experiment with detection limits?

I would think that a pulsed pressure injection would be needed to inject 12.5ul at that temperature without causing backflash of the solvent.

[Rndirk]

Well it seems they prevent backflash somehow. I am interested how this works:

A fast autosampler injection with liquid band formation into a liner containing glass wool is used to prevent backflash in the injection port [3].
This injection method produces a pressure surge from the vaporizing solvent, forcing the analytes onto the pre-column

On the other hand, why don't you go for solvent vent since you're capable of doing that? 12.5µl sounds like a lot to condense on a pre column, even with 0.53 diameter.

[James_Ball]

I will be looking at solvent vent once we can get the Agilent Multi Mode Inlet, but I saw this as something to try in the interim. I could cool the normal injection port before each run, but I don't think it will heat fast enough do to it properly.

I wonder if putting the gooseneck at the top would also help prevent backflash?

[markf]

Sounds interesting, a couple of questions:

1)How much actual volume goes into the MS?
2)What method are you using for semi-volatiles, what matrices are you usually analysing?

I'm asking because we work with EPA8270D and we inject 1uL with 0.5:1 split, and analysing both USEx extracts from solids, LLEx extracts from wastewater and direct injections (diluted) of organic wastes, and from my experience, our sVOC analysis is the reason we need to clean our source more often than not.

[James_Ball]

We will be working with EPA 625 and 8270, then moving on to 608, 8081,8082 for pesticides and PCBs.

With the Restek method most is going on column, once we can get the Multi Mode inlet, most should be vented out using the solvent vent technique.

We run almost every type of sample, from waste dilutions to drinking waters to soils.

The main push is to reduce solvent usage, and after reviewing the methods, EPA3510C for sep funnel extraction allows to reduce initial volume to 100ml and you use the same 90% reduction in extraction solvent. Method 625.1 was modified so that the initial volume of sample needs to be enough to meet MDL and reporting limit goals and the final extract volume has the same stipulation.

As for needing to clean the inlets, by extracting 100ml sample and injecting 10ul extract you should get the same amount of analytes or crud into the injection port as you would by extracting 1000ml and injecting 1ul. There will probably be some increase in maintenance needed, but if we can achieve a 90% reduction in the methylene chloride it should more than make up for extra liners and gold seals.

I may also experiment with it on the drinking water side to allow for no solvent concentration after SPE, which will save time also on those, if I can dilute up to 10ml instead of concentrating down to 1ml then injecting 10ul. Since that should be fairly clean matrix, it should help with those also. Not really much difference in solvent usage, but that much less that goes up the hood and into the air.

[Rndirk]

I'm curious: do the EPA methods allow you to take a subsample of 100ml water from a 1L sample bottle? Do they specify guidelines to do this?

When you take 100ml, you can't rinse the empty bottle with extraction solvent.. Apolar stuff like PAHs & PCBs can stick to the glass or particles in the water.

[James_Ball]

Subsamples are not allowed by the EPA methods for the reason you pointed out, you must rinse the sample bottle with solvent. They do however allow for usage of smaller bottles, which in turn will reduce shipping costs due to reduced weight of sample.

[mckrause]

James,

I run CSR all the time. Mostly what I do now is pesticides in food, so primarily I'm dealing with acetonitrile. However, I was using it for 625/8270 routinely a couple of years ago. I love it - with methylene chloride we were routinely injecting 10 uL with no issues at all. I never saw the fast injection issues that Jack Cochran documented, but I am using a Leap autosampler, not the Agilent.

Sure is nice not having to concentrate dirty samples down to 1 mL or less. I highly recommend you try it.

[James_Ball]

I am experimenting with it now, using a 0.25 guard column until my 0.53 arrives. So far things look good. I did add a step in the oven program that holds at 35c for 1 minute to condense the MeCl then ramps 30C/min to 45 and holds another minute to allow the solvent to evaporate off, then I start the normal oven program.

Experimenting now with optimizing the pressure pulse and splitless timing to control a small amount of peak tailing I am seeing on the later eluting compunds. Great sensitivity, I am definitely going to start using it on the 525.3 Method I am working on.

[sav]

Has anyone tried the LVI procedure listed in this tech note?

I tried this CSR LVI approach and it worked fine up to 20 - 30 ul (heptane, hexane), but at 40 - 50 ul (heptane, hexane) i found some recovery problems (instability). In general, 0.53 mm ID retention gap is much better (recovery and max injection volume) than 0.25 mm and initial oven temp must be much lower then boiling point of solvent (so solvents with higher boiling point are preferable).

[James_Ball]

Thanks for the input. I am glad to see others have had success with this procedure too.

I would like to go with Ethyl Acetate as my solvent but I don't think EPA will allow the change, and it could interfere with the early compounds, Pyridine being the first to elute. Going to the multi mode inlet and using CO2 would probably be even better but for now this is a much less expensive route and looks like it will achieve our current goals of reducing solvent usage.

I just installed the 0.53 guard column this morning, once I get the press tight conditioned I will make an injection and see how it goes, I am hoping for some really good results. Once I get it working on the semivolatile mass spec methods I will start looking into using it for the Pest/PCB on the GC/ECD tests.

[sav]

I would like to go with Ethyl Acetate as my solvent but I don't think EPA will allow the change, and it could interfere with the early compounds, Pyridine being the first to elute.

I think its (huge solvent band) a main disadvantage of this CRS LVI method, especially for cases when solvent and analyte have relatively close RI. I think AcOEt will interfere (CSR LVI) with Pyridine on non-polar and semi-polar phases (even on polar phases), so better chose another solvent.

Once I get it working on the semivolatile mass spec methods I will start looking into using it for the Pest/PCB on the GC/ECD tests.

I think it will give some benefits if you want do some optimization of protocols/methods (reduce sample weight, solvents consumption etc.), but if you want increase sensitivity by this simple way - i think it will not work fine without improving extract clean-up step.

[James_Ball]

I would like to go with Ethyl Acetate as my solvent but I don't think EPA will allow the change, and it could interfere with the early compounds, Pyridine being the first to elute.

I think its (huge solvent band) a main disadvantage of this CRS LVI method, especially for cases when solvent and analyte have relatively close RI. I think AcOEt will interfere (CSR LVI) with Pyridine on non-polar and semi-polar phases (even on polar phases), so better chose another solvent.

Once I get it working on the semivolatile mass spec methods I will start looking into using it for the Pest/PCB on the GC/ECD tests.

I think it will give some benefits if you want do some optimization of protocols/methods (reduce sample weight, solvents consumption etc.), but if you want increase sensitivity by this simple way - i think it will not work fine without improving extract clean-up step.

We are looking mostly at the reduction of solvents by using a proportional reduction of sample. 100ml sample instead of 1000ml and 18ml solvent instead of 180ml. I know of other labs doing this on a routine basis, we are just now trying to do the shift to less solvent usage so it is a bit of a learning curve.

The only problem I am running into now is with tailing of Pentachlorophenol and Benzidine.

It is a problem that seems to be carrying through from our normal 1ul injections where we quickly lose response for Pentachlorophenol just when making the tune check injections. When I first ran semivolatiles 20 years ago I could go days without inlet maintenance and not lose Pentachlorophenol, and once it did begin to drop off simply changing the liner and removing 5-10cm of column would bring it back. Now we have the problem that we lose Pentachlorophenol within a few injections. We even have an instrument where we replaced the injection port with a new one, changed the column and liner and after a few injections it began to drop off and soon was gone completely. But if you install a Uniliner the pentachlorophenol returns. That would make it seem to be a problem with the metal injection port, but I have never seen one become active/adsorptive from just injecting standards. I am beginning to suspect our Methylene Chloride has something to do with it. Since we moved away from using Fisher Optima grade solvent that may be the time this began, but I am not sure what in the solvent would cause a problem like this. Has anyone else had a problem with rapid deterioration of the injection port before?

[Bigbear]

I had some ethyl acetete that had something in it that caused endrin/DDT breakdown to go >20% after 10-12 injections. Can't remember what the library search claimed it was.
As a result we now have a policy that we have to blow down 20-25 ml of a solvent used in 525 to 1 ml and shoot it. It can't have any significant peaks in it.

[James_Ball]

We were doing solvent lot checks here in the past too, but since I have been back over in Volatiles a few years, it seems that fell by the wayside. Also with the ones above making decisions to lower expenses, I am not sure now if we still have the best quality supplies. Now that we are going to be using LVI it is going to matter that much more that we have high quality solvents and other supplies too, so I guess I will be doing a review on every part in the chain as I develop the methods.