Chromatography Forum

Hello!
I am working with GC gas natural analysis with a 6890 GC, I have some problems because there are 4 "valves peaks" so they given peak coelution with CO2 and C1 in TCD, then the restrictor doen´t works well because it lets that those peaks appears, I read about some columns that can replace the restrictor but...What kind of column Can I use ?

Thanks very much!

The best packed columns 'buffer or restrictor' are those that are fused silica or glass coated SS with minimum deadspace (packed fully with screen terminations) packed with an inert support such as Chrom 750 or Supelcoport. Don't use wire terminations as they can become blocked, full or partially, which changes the backpressure and flow rate of your 'train'. Minimize your system deadspace for best results.

best wishes,

Rod

What ever you put into the your GC, restrictors or needle valves, will give you 'peaks' when you switch the valve using a TCD. You should first try resetting the needle valves, as your columns may have 'aged' and the restriction of your column has changed. This should reduce the effect. The setting up of the needle valves is given in the manual supplied with the analyzer. You have a system with multiple columns in series, but some columns have to be switched out of stream to prevent compounds being absorbed completely. This is especially true for the Molecular Sieve column, which is the last column in the series. During normal operation, the exit of this column is at atmospheric pressure. When the column is isolated, the pressure in the column will equilibrate so that it is the same throughout the column. When you switch the column back into the stream, the exit of teh column is no longer at atmospheric pressure, so you will get a flow surge which will be seen as a baseline change on the TCD detector.

You could try adding a buffer column in front of the TCD as suggested by chromatographer1, this should reduce the surge. Replacing the needle valve with a fixed restrictor will not be easy as it is difficult to know the actual restriction of the column under the conditions that you are using it.

Gasman

You can also use a 1/16" OD piece of tubing with 0.010" ID or less. You simply cut the tube shorter and shorter until you have the proper restriction you desire.

You can use this as a buffer for the surge upset when a reverse column step or backflush to detector configuration is used, or as a parallel column when used in a trap. Be sure to check flows only when thermal equilibration has been reached.

What setup (valve column configuration) are you using? Where are you wishing to put a buffer column?

Where are the valves which don't seem to be working for you? What you describe can also be a sign of a slight leak in the 'train'.

I have extensive experience in this type of analysis. (I developed a 3 min C6+ BTU application) and a 2 min application with an additional valve)

good luck,

Rod

Hi Gasman & Rod!
Thank you very much for answer to my question.
I'll try with an old DB-1 column that I have, I hope it work.
I have three columns for TCD: one make the backflash of C3+, another separate CO2 and C2 ( + H2S peak ), I would like to know if Molecular sieve column besides separate O2, N2 & C1 can work with CO2?, because in all papers and application notes don`t link a molecularsieve column with this gas (CO2).
I didn`t move the time valves switching but some weeks ago I dind`t find de H2S peak of my reference gas What could happen?.

Thanks again.
Best regards
Juliana

Most molecular sieve columns will not elute CO2 as a symmetrical peak due to the manufacturing processes used to make the column.

It can be done with an excessive amount of temperature and flow however. You will not wish to try it though unless you prepared the column yourself and you know what you are doing. I believe I have posted chromatograms showing this elution previously on this forum; search the archives.

H2S reacts with metal active sites. It does not elute from Mole sieve columns easily.

Any buffer metal tubing without a glass or fused silica coating will react with H2S.

You don't want the CO2 or C2s or H2S to get on the mole sieve column.

It is easier to set valve timings when you have a greater separation of the CO2 and methane peaks. Packings like porous polymer N, S, R, or A are to be preferred over Q, P, or D.

You could have set the timings or changed the columns in your analyzer where the H2S peak did not elute off the porous polymer column where it could bypass the MS column. Then in the next run it might have eluted off onto the MS column and was trapped there.

Good luck with your work,

Rod

Something like chromatographer1 say, you can use a 1/16 tubing and blend it.

Thanks Rod & Gamarra .

I´ll tell you what happens .

Best regards!

Hi boys!

After try all I couldn't remove the valve peaks, so I`ll have to work with it.

Now I have another question: I'll report C1 y C2 from TCD and C3+ from FID, It will be necessary to do some kind of "tied" between the response from both detectors or only I have to get the response factor for each component in each detector, multiply each area by that response factor and to amount all it fot the normalization?


Thank you very much

Best regards,

Juliana

You do not have a simple measurement.

You are not able to tie together the responses from each detector and normalize.

You can only normalize when you know in full and exactly the content of your matrix.

This you do not know, you can only assume. You will not be able to state precisely the amounts.

You can prepare mole standards for the gases you are measuring and calibrate a single point calibration and calculate the peak areas from this calibration.

If you assume perfect linearity of response and mole mass over the range you are measuring you can estimate the matrix contents.

But a 3+ peak will always be an estimation. You have no way of determining the actual content and response of this peak.

I hope this makes sense to you. Your expectations are high. Unfortunately, the reality is cannot normalize your results.

Rod