Running a TCD and FID in series

[GHGirl]

I have a fairly new Agilent 7890A GC. Its internal configuration is becoming complicated: a capillary CarbonPLOT column connected to a Deans Switch, with one switch routing towards a TCD and the other to an uECD (I'm using .10 mm x 1 m restrictors). However, I have one more detector, an FID, that I would also like to use.

Has anyone had experience connecting a TCD to an FID to run them in series? It looks like I'll need a 1/8" to 1/16" Swagelok reducer to make the TCD vent to restrictor connection... Is there a kit available for this? Is this a bad idea? Suggestions?

My last resort is to use a "Y" press-fit union to split the sample between the TCD and FID, but I'd like to get the full sample to each of the detectors, if possible.

Thanks for your help,
Greenhouse Gas Girl

[chromatographer1]

If you are using a TCD like I had in my 6890 simply connect the TCD vent (a 1/8 inch tube) to the FID using SS reducers of 1/8" to 1/16" or to 1/32", use graphite ferrules and 0.32 or 0.53mm FSOT tubing to make the transfer connection non-permanent. If you are measuring anything that will condense at room temperature then insulate the tubing.

The faster the flow through the TCD the less a problem you will have with peak broadening in the void space before the connection of the 0.32 (or 0.53mm) FSOT tubing into your FID. I would NOT use anything less than a 0.32mm ID FSOT tube (make it as short as possible) as that will increase the back pressure on the TCD and increase the peak broadening issue.

A couple of graphite ferrules and a reducer union (Swagelock or Valco) should take care of your problem.

best wishes,

Rod

[GasMan]

GHGirl,

There is a kit that will allow you to connect the TCD in series with the FID. The part number is 19232C. However, you need to have the correct FID installed. There are two flavors of FID for the 7890 GC. The first is the FID which can ONLY be used with capillary columns. The second is the FID which can be used with capillary columns or packed columns, depending on the adapter that install into the base of the FID. The 19232C kit requires that you have the FID that can be used with packed columns. If you look at the base of the FID that protrudes into the oven, you can determine what you have installed. The packed version has a 1/4 inch fitting which will take adapters for 1/8 inch packed columns or capillary columns.

If you install this kit, the exit line of the TCD is brought out into the oven.

Gasman

[GHGirl]

I thought I'd post a quick update for others who may run into the same issue in the future.

I've been working closely with Agilent to find a solution to running three detectors simultaneously on our 7890A Gas Chromatograph. One of the major challenges is that our particular system is capillary, including the FID.

As Gasman mentioned, the kit that Agilent offers for running a TCD and FID in series is only for a packed FID, so that wasn't an option for us. But we did manage to clip the 1/8" flare off the TCD vent and connect the 1/16" portion of the TCD vent to a 0.53mm capillary using a Swagelok union, which could then be inserted into the FID (as Rod suggested). In fact, we tried several variations of this connection...

Ultimately, the TCD mechanics (switching from the carrier to a reference) was a problem for the FID baseline, and we could never get the noise quiet enough to detect trace amounts (i.e., atmospheric levels) of methane on the FID. So for now, running the TCD and FID in serious on a pure capillary 7890A system does not seem to be an option.

Our solution to running three detectors simultaneously? We connected the column (we switched to an HP-PLOT/Q) to a Deans Switch, with the OFF position running to a Two-Way Split and the TCD and FID (so the sample is split in half), and the ON position running directly the the ECD. In this way, we can time the Deans Switch so our CH4 and CO2 reach the TCD and FID, and the N2O reaches the ECD. We can also run C2H2 (see my post, Acetylene Poisoning ECD - Fact or Fiction?) to the TCD and FID, regardless of whether C2H2 is a true problem for ECDs.

As always, thanks for help!
Greenhouse Gas Girl