Area Reproducibility & Other MISC Horrors...

[Peter Apps]

Hi July

Although the split ratio is set on the GC and is set at the same value for each injection, the actual split ratio that you get with each injection depends on the gas flow into the column and the gas flow out of the split while there is sample in the inlet. The flow and pressure controllers on the GC inlet have a limited ability to maintain pressure and flows at the set values when the flow and pressure suddenly increase and decrease as you inject the sample. Disruptions are minimised by smaller sample volumes (which give smaller peaks) and slower injections (which give wider peaks).

In principal an autinjector will give better repeatability, because it gives consistent injection speeds, but off hand I cannot think of one that will be able to take the sample from the 1 l gas bulb, although I suppose you could rig a CombiPal, MPS2 clone type by setting the syringe positions specifically for this job.

An alternative to syringe injections would be to use a 6-port valve, but you are still not going to be able to get down to the necessary detection limits in real samples.

The problem boils down to this; at the concentration levels that are relevant, the volume of sample that contains enough analyte for the detector to consistently measure is larger than you can put onto a capillary column without causing volume overloading (broad peaks). Therefore you have to reduce the volume of the sample without reducing the quantity of the analytes. You cannot do this by cold trapping because real samples contain water vapour, and so sorbent traps are the only practical way of getting the job done. Yes they have drawbacks, but these are more tractable than the problems with alternative approaches - some of which you have already experienced.

Good luck

Peter

[july]

Thanks for all of your help. We do have a Tekmar-Dohrmann Purge and Trap Concentrator 3100, but to be honest I'm still overwhelmed at the moment ...still a little bumfuzzled as to how this crazy stuff works. If I am reading the concentrator manual correctly, it is only for soil or water samples? It came with our GC/MS...no one has ever used it. I was comparing the water removal techniques mentioned in TO-14a and TO-15. It seems both methods have identified errors associated with them. Does anyone have an opinion on which is easiest or less costly to set up or which is the most accurate/reliable?

TO-15:
To analyze the sample, a known volume of sample is directed from the canister through a solid multisorbent concentrator. A portion of the water vapor in the sample breaks through the concentrator during sampling, to a degree depending on the multisorbent composition, duration of sampling, and other factors. Water content of the sample can be further reduced by dry purging the concentrator with helium while retaining target compounds. After the concentration and drying steps are completed, the VOCs are thermally desorbed, entrained in a carrier gas stream, and then focused in a small volume by trapping on a reduced temperature trap or small volume multisorbent trap. The sample is then released by thermal desorption and carried onto a gas chromatographic column for separation.
As a simple alternative to the multisorbent/dry purge water management technique, the amount of water vapor in the sample can be reduced below any threshold for affecting the proper operation of the analytical system by reducing the sample size. For example, a small sample can be concentrated on a cold trap and released directly to the gas chromatographic column. The reduction in sample volume may require an enhancement of detector sensitivity. Other water management approaches are also acceptable as long as their use does not compromise the attainment of the performance criteria listed in Section 11. A listing of some commercial water management systems is provided in Appendix A. One of the alternative ways to dry the sample is to separate VOCs from condensate on a low temperature trap by heating and purging the trap.



TO-14a:
The Compendium Method TO-14A analytical system employs a Nafion® permeable membrane dryer to remove water vapor from the sample stream. Polar organic compounds permeate this membrane in a manner
similar to water vapor and rearrangements can occur in some hydrocarbons due to the acid nature of the dryer. (Uses cryogenic preconcentration somehow it would seem)?

Thanks for your time.

July

[july]

Although the split ratio is set on the GC and is set at the same value for each injection, the actual split ratio that you get with each injection depends on the gas flow into the column and the gas flow out of the split while there is sample in the inlet. The flow and pressure controllers on the GC inlet have a limited ability to maintain pressure and flows at the set values when the flow and pressure suddenly increase and decrease as you inject the sample. Disruptions are minimised by smaller sample volumes (which give smaller peaks) and slower injections (which give wider peaks).

Ok. When we were using a 10uL injection a GC/MS tech support guy suggested that we use a larger sample to reduce compressions of the sample in the syringe, so we increased the injection to 100uL. I'm not sure which is the lesser evil - pressure changes in the GC or compression of sample in the syringe.

[JI2002]

If you let us know what you want to achieve, I probably can offer some suggestions.

Here are some of the questions coming to mind:

What are the target compounds?
What are the reporting limits for those compounds?
Will the samples be collected in canisters, tedlar bags or 1L glass bulbs?
Do you have some kind of autosampler with Tekmar 3100?

Tekmar 3100 is not that hard to operate, you can call technique support to get it going. In order to run air samples, you can either run it in soil mode with Archon, (You can only use this for high concentration samples) or you can buy an autosampler just for tubes. I'm sure Tekmar can give you more advice.

As far as TO-14A or TO-15, you need a lot more hardware to run these methods. It's not easy, it's tricky, sometimes confusing. That's why there are so many environmental labs in US, only a small portion of them offer these analysis.

[Peter Apps]

Hi July

The sample in the syringe compresses as soon as the needle penetrates the septum becuase the inlet pressure is above atmospheric. The amount of compression depends only on the diference between the pressure in the syringe (atmospheric) and the pressure in the inlet. It is completely independent of the syringe volume.

Peter

[Bruce Hamilton]

The reason TO-14A and TO15 are complicated is because quantitation of trace volatiles in ambient is difficult. If you have only a couple of gases to determine, the methods could be simplified, but you still have to calibrate your sampling system.

You need accurate standards with a sampling system amenable to trace gas analysis, including accuarte sample volume and/or pressure. Benzene and toluene are also significantly ( > 0.1% ) soluble in water, so water removal has to be specific.

One relatively-cheap method for cyrogenic trapping of trace organics in air is a simple short packed GC column that is immersed into a cold bath in a Thermos flask whilst sampling, then the Thermos quickly removed and replaced with another Thermos filled with hot water, and using a six-port valve to switch the trap into the column. The advantage is that you can select the packing to match the compounds.

Benzene and toluene are pervasive, it's easy to get background from all sorts of polymers, so you have to be diligent removing polymeric lines etc. and ensure that you don't contaminate you systems with high standards.

There are also plenty of published research methods, and you should look for ones that use similar equipment to what you have available.

Bruce Hamilton

[july]

If you let us know what you want to achieve, I probably can offer some suggestions.

Here are some of the questions coming to mind:

What are the target compounds?
What are the reporting limits for those compounds?
Will the samples be collected in canisters, tedlar bags or 1L glass bulbs?
Do you have some kind of autosampler with Tekmar 3100?

Tekmar 3100 is not that hard to operate, you can call technique support to get it going. In order to run air samples, you can either run it in soil mode with Archon, (You can only use this for high concentration samples) or you can buy an autosampler just for tubes. I'm sure Tekmar can give you more advice.

I had been given a list of target compounds, but I checked back with the chemist to double check, but I think I misunderstood. What I was told was that we are going to choose a compound to represent each of several different types of chemical compounds. For example, he's decided to use toluene to represent aromatic/benzenelike compounds, d-limonene to represent alkene compounds, acetone to represent ketone compounds. He has not decided on a substance to represent aldehydes, esters, or straight chain hydrocarbons as of yet because we have not made a final decision.

I work at an air conditioning company. The filter group would like to be able to measure the change in VOC levels in the air before and after filtering to see what the filter is actually "cleaning" out of the air.

Originally, I was told they would like to see contaminants on the order of 10 ppb within an error no greater than 10% when we inject the air samples.

So ideally, we would like to use the 1L glass bulb for our sample unless there is some reason that we can not.

We do not have an autosampler, but we are considering buying one if needed.

But before we buy the gas standard...there's no way that we can "test" the GC/MS out to make sure that it is even capable of doing what we've planned is there? Not without buying the gas standards?


Hopefully I answered all of your questions.

Thanks again all.

[JI2002]

First, if the GC/MS functions properly, it definetly can handle the analysis, no matter what kind of concentrator you are going to use. By doing TO 14A/TO15, you can have a reporting limit of 0.1 ppbv for almost all the VOCs. Reporting limits are a liitle bit higher for TO17 at about 1 to 2 ppbv depending the sample volume in the tube.

Second, hold off buying any gas standard until you have a better idea what compounds you are going to do. A TO 15 mixture @ 1 ppmv(1800 psi) standard can cost a few thousand bucks, and it may not have the compounds you need.

10 ppbv detection limit sounds a little high, but if this is really what they want to do, it's possible for you to do it without a concentrator. I don't have experience of doing it, but off the top of my head, if you can use a cryo focusor(in liquid N2) to trap 10 ml sample, and flush everything(no split) into the GC column with a carrier gas flow of about 1 ml/min, you should be able to see 10 ppbv without a problem. Moisture won't be a concern for sample volume of 10ml.

[JI2002]

To finish up what I left out, here is what you can do to make it work:

Install a deactivated metal tubing to the split/spitless injector port (set the injector to run in split mode)just as you install a column, run the tubing to a six-port valve. Connect another tubing to the six port valve, run the second tubing back to the column via a zero volume union. Part of the second tubing will be used as cryofocusor. (the rest of the tubing should be heated at about 100C). Then you need to install a 5 ml sample loop to the six port valve.

Here is how it works: When the GC/MS is ready, cool the cryofocusor (part of the second tubing) to -160C in liquid N2, then load the sample to the 5 ml sample loop(flush the loop with the sample first), then turn the six port valve, the carrier gas flow transfers the sample to the cryofocusor. N2 and O2 will pass through the cryofocusor to the column, the rest including VOCs, water and CO2 will be trapped inside the tubing.
When the transfer is finished, start GC and warm up the cryofocusor to about 100C in a very short period of time(a few seconds), VOCs will be transferred to the head of the column in tight slut. The rest should be easy.

This is similar to Bruce's idea, the difference is that no trapping materials is involved in this approach. There are commercial instruments available for this application. It's possible to do it mannually, but the results can be compromised.

This method is only good for high concentration samples due to its low sample volume, but it shouldn't be a problem to meet the reporting limit of 10 ppbv.

Although a split/splitless injector is used in split mode, this is a splitless injection because all the target compounds are tranferred into the column.

[july]

There are commercial instruments available for this application. It's possible to do it mannually, but the results can be compromised.

First of all...thanks everyone for all of your help. I appreciate it very much.

Secondly, I have a few more silly questions:

What are these commerical instruments called that do this application?

Not using the commercial instruments can compromise the results?

Thanks again.

[JI2002]

7032 autosampler from Entech(simi valley,CA 815-527-5939) is the one we use. You can call other manufactures to find out which works best for you. Here are some of the companies that may have similar autosamplers: Tekmar, Varian, XonTech, Graseby, Dynatherm, Nutech.

The autosampler is controlled by software, so it's all automatic. The results are consistent. If you do it using homemade setup, it's going to be hard for you to meet the accuracy criteria of 90 - 110%.

In case you change your mind and decide to send samples to a commercial lab, give us a call(just kidding)

[july]

In case you change your mind and decide to send samples to a commercial lab, give us a call(just kidding)

lol..Thanks! I don't even know what lab you are genius;)

[AICMM]

Several comments about what I read into this posting. First, I have done gas sampling bulbs many times and gotten very good results. You do need to do mixing in the bulb and you do need to keep the bulb warm. Warm, you can use any GC oven. Mixing is simple. Take about 5 or 6 glass beads and drop them into the the glass sampling bulb. Then when you are ready to take an injection, roll the gas sampling bulb around in your hands so the beads rattle around in the bulb. This was the technique used to make a large number of standards for environmental monitoring in one of my previous jobs. Very good reproducability. Second, if you are putting neat materials (acetone or whatever) into the gas sampling bulb, even it is only a few microliters, you should be injecting plenty of material on column assuming a reasonable split ratio. Third, you can investigate propane cans as a means of preparing standards as well and you get much more standard for your efforts. If you would like I can tell you an easy way to do that.

[july]

I would LOVE to hear. Do tell!

So rather than injecting the liquid into the glass bulb, you prepared in the propane container and transferred to the bulb?

Secondly, what sample size and split ratio would be reasonable using this method?

Thanks!

[AICMM]

July,

If you make a propane can standard, you simply put the solvent into an evacuated propane can and then back fill with air or balance gas. If you do that, you can run the can directly and the sampling bulb goes away completely. You will end up with ppb to low ppm values based on how much material you put into the can and how much balance gas you add. Not necessarily accurate (although I have had good luck so far) but it gives you a bunch of standard so it can be very precise. At ppb levels, a syringe injection or a loop injection is going to have a hard time seeing the analyte. I do it all the time but I don't use an MS I use my own detectors and ppb in air is pretty easy. So far, I have not seen anything on your actual split ratio and that is pretty critical to defining your ultimate detection limit. Also, my experience has been that trying to dilute via split ratio did not work well so I hope you are not doing that. I would second the earlier comment that a gas sampling valve would be a better way to make these injections since it avoids pressurising the injector when making a syringe injection and you get much better peak shape. Finally, a 3100 can most certainly be made to do your concentrator work but it requires more discussion than I can easily type out.

Feel free to call me if you wish to discuss any of these things further:
Matthew 505 890 6096.