Bypassing column and inlet to use FID

[aaron]

Looking all over the web, this looked like a good place for someone who knows next to nothing about chromatography to ask a question.

My lab is trying to test the output of an olfactometer (used to present odors to subjects) and thus far we have been somewhat unsuccessful in our attempts to do so with GC. We've tried sampling the output stream with a syringe as well as feeding the output directly into the split/splitless inlet on our Agilent 6890 via a sampling valve we scavenged from a LC setup, but neither technique works very well. The first is problematic because it requires the olfactometer's output to be "on" for a longer time period than it will be in practice and the second technique has more problems than I should even consider going into here.

My question is this: how feasible would it be to hook up the output of our olfactometer directly to the FID, bypassing all the other components? It should be noted that we are working almost exclusively with single component analytes in our output stream. These range from aliphatic acids to alcohols to aromatic aldehydes. The total output stream of the olfactometer ranges from 100-1000ml/min, but in actual use, it will probably be somewhere between 300-500ml/min, so we probably want to stay in that range when sampling. I suppose my biggest concerns are:
a) How do we physically hook it up? What kind of adapters are available to interface the FID inlet with larger tubing diameters?
b) Can we use air, which is our normal carrier medium for the olfactometer, or do we need to use helium or nitrogen since we are trying to feed into the FID?
c) Would we need to install a different size jet than the one which is standard into the FID in order to use flow rates as high as our output and/or do we need to bleed off part of our output so that we have a smaller amount actually going into the FID?
d) Will this damage our FID or other components of our GC in any way?

Thanks for your help, and as I said, I am a complete beginner when it comes to GC stuff, so feel free to be as explicit as possible when explaining things. Also, for reference, please see:


http://www.jgp.org/cgi/reprint/74/1/37. ... r%20FID%22
and
http://www.sepsci.com/chromforum/viewtopic.php?t=2797

Thanks in advance for your help!
-Aaron

[Consumer Products Guy]

At first glance I don't see why you couldn't use your olfactory air stream of 300 ml/min in place of the traditional 300 ml air delivered to the FID.

[aaron]

That seems like it could be a great idea, but I am worried that it will be more difficult to hook it up this way. If I attach where the column normally attaches, I thought it would be easier than disassembling the other connections to the detector, but I really have no idea since I have never done it. Any thoughts on that Consumer Products Guy? And by the way, thanks for your quick response.

[GasMan]

Aaron,

Before you connect up to the air supply as suggested, you need to think how the FID works. It depends on the sample being ionized in the flame, and these ions are collected as the signal. Not all of the air is used in the combustion process, some of it will be used to sweep the combustion products out of the detector. This would mean that if your air contained the 'sample', not all of your sample would be burnt, and therefore the signal would NOT be a true representation of your sample.

I believe that you will need to use a different gas, such as nitrogen, and feed this into the bottom of the FID. However, a flow rate of 300 ml/min, will also give your problems with the standard jets supplied with the FID. If the velocity of the gas is too high, the flame will not stay lit. You may need to drill out a standard jet so that the velocity of the gas leaving the jet is lower than the flame speed for a hydrogen flame.

By the way, I tried your first link, but have had no success.

Gasman

[pi3832]

It might be easier just to buy a used Rosemount model 400A (a.k.a. Beckman model 400A). That's pretty much what they are: a stand-alone FID.

[akthmps]

Are you just trying to determine what it is the output? If so, run the output stream into a flask, and make sure you have another section of tubing coming out of the flask. Then insert a SPME through the stopper and collect a sample. Then I'd use GC-MS to characterize the components.

I really didn't read carefully through your first post, so maybe this isn't of interest, but just thought I'd throw in my two cents worth...

[Peter Apps]

Hi Aaron

I am assuming that you need the signal from the FID to compare between dilutions on the olfactometer. If you need to quantify (in terms of mg/cubic m or whatever) then things get more complicated.

Assuming that the output of the olfactometer is well mixed (and if it is not the olfactometer is not working right !) you do not need to feed the whole of the air stream into the FID. Using suitably inert hardware, connect a T-piece to the olfactometer outlet. Take a tube from one leg of the T towards the FID but do not connect it yet. Take a tube from the other leg of the T to a needle valve. Connect a bubble flow meter to the tube that will be going to the FID, and udjust the needle valve until you get 20 - 30 ml a min through the bubble flow meter. Connect the FID tube to the FID, and run the FID with no make up gas.

This set up will give some back pressure at the olfactometer outlet, whether this is a problem will depend on its internal plumbing.

If you change the olfactometer flow rate you will need to re-adjust the needle valve to maintain the same flow into the FID.

This is pretty crude, but it might do the trick. Good luck.

Peter

[aaron]

GasMan-Incomplete combustion was exactly the problem I was worried about with using our olfactometer output as the air supply to the FID; thanks for confirming this. As far as using Nitrogen, is there a particular reason to do so? Also, for the jet, do you think one of the larger bore jets available for use with packed rather than capillary columns could handle larger flows?

pi3832-Any idea how much a Rosemont 400A can be come by used? Also, how hard would it be to find a used model with the optional current output board to hook up to a resistor and use with an ADC to take continuous measurements?

akthmps-When you say "are you just trying to determine with it is the output?" do you mean what are the components or what are the concentrations? We will know the components, we are trying to determine the concentration. This leads me to think that SPME may not work well because some of the sample could be lost through the outlet without adsorbing onto the SPME fibers. That said, I suppose it would be possible to somehow mount the SPME sampler directly in the outlet stream, forcing all the output gas over the fibers, but I really have no idea how that could be done. We considered using activated carbon sampling vials, but the problem with doing so, and with using SPME, is that it makes it quite difficult to get accurate time course measurements.

Peter Apps-This idea seems somwhat feasible, but I have a couple questions about it. First, not knowing all that much about bubble flow meters, do they pass all the flow through the oil/water/whatever is being bubbled through or just a small portion of it? I would have some concerns if all our flow were going through solution before going into the FID. Second, how would we want to go about connecting between the flow meter and the FID inlet? I assume we would need some sort of adapter, but the only FID inlet types I have seen are designed for columns. Finally, we may be able to scavenge a digital flowmeter from one of our other olfactometers to temporarily make measurements on flow to the FID, do you think this or a bubble flow meter would give better results. For reference, we would be using something like an Aalborg GFCS 010312, which combines flow control with flow measurement, which would emiminate the need for a needle valve (possibly).

All-Thanks for your great responses, and please continue the input. It is appreciated.

-Aaron

[Peter Apps]

Hi Aaron

You would set the flow with the bubble flow meter (or a fancy electronic equivalent) at the end of the line that will go to the FID, then you disconnect the meter and connect the line to the FID. Because you have 20 - 30 ml/min through the line you can connect direct to the base of the FID - you do need a line thin enough to thread up inside the stem.

Reading again I start to wonder if you want to monitor concentrations while a subject is sniffing, as opposed to using the FID just to check the olfactometer performance in advance. If this is what you want it will be a bit more complicated to arrange, but nothing is impossible.

Peter

[aaron]

We don't really want to check with a subject and FID simultaneously. Disconnecting/reconnecting should work fine. Thanks for your advice.

[christiancarioca]

I am currently using a FID for measuring methane (from reduced COx in an external, home-made methanizer) as a continuous detector. The only modification I needed was to install a three-way valve in the H2 combustion line and then turn off the N2 makeup. Air is supplied in an independent line. Instead of combustion H2, a mixture of N2 (37 ml/min), H2 (38 ml/min) and Ch4 (max 1%) arrives at the detector. It works great!!! And it is very sensitive.

[GasMan]

Aaron,

The reason that I mentioned nitrogen, is that you will need an inert gas mixture going out of the jet. if you have any amount of air present in this stream with hydrogen, then it would be possible for the flame to burn INSIDE the jet. It is possible to get a 0.030 inch jet for your 6890. It has the part number 18789-80070.

Gasman

[aaron]

christiancarioca--thanks for your idea, we might try it.

GasMan--Good to know, I hadn't thought about burning inside the jet. We will have to use something inert (probably helium, but maybe nitrogen). Also, we rather unfortunately have the capillary optimized FID, so we can't use the 18789-80070 (I think). Instead, we are getting a G1531-80260, which is a 0.018in jet made to fit the FID parts we have. If you know that the non-capillary optimized jets can be used with our setup, let me know.

Thanks for both your responses,
Aaron