[Wayne Bridgeman]

Hello all:

I work in a ME lab and do not have a Chem. E. background. Our center is looking to build a bench scale mild gasifier. The output of this plant will be what can broadly be called a "syngas". We will want to analyze this gas. The gas will have lots of long and big hydrocarbons in it (coal tar) that present complications in analysis. I have explored many options through research and contact with equipment vendors including FTIR, EDX, GC/MS, GC/FID-TC. I have really settled on the GC/FID-TC as the best solution. We are looking for Vol. % of H2, CO, CO2, H2O, N2, O2 and a great big lump of hydrocarbons from C1 up to C22 and possibly higher. Because of the heavy hydrocarbons the gas needs to be kept above 524*C. This is one of our largest complications.

We originally wanted a general formula (ratio of H to C) for these hydrocarbons but could not find a solution. So have resigned ourselves to just getting the volume percent of the lump of hydrocarbons.

I ask for your thoughts. Have I overlooked an obvious solution? Can an empirical formula for a broad range of hydrocarbons be determined? How?

Thanks.

   

[marquis]

I'm familiar with GC/FID, but not GC/FID tc.  Could you please explain that method?

You have a wide range of substances you are trying to analyze.  That seems to be the major issue.  It is relatively easy to identify nitrogen, oxygen, and other low molecular weight gases using a molecular seive column and a thermal conductivity detector (TCD).  This setup does not work as well for higher molecular weight compounds.

We used GC/MS (with an HP5 30 m microbore capillary column) for identification of the different molecules in oil and wax samples.  It covered many of the same range of carbon numbers you are talking about.  You could inject a standard known molecule and record its retention time and mass spectrum.  The mass spectrum could be input into the instrument library for future identification.  Having said that, many of the hydrocarbons are hard to differentiate unless you can find the molecular ion.  In that case, retention time would be needed as well.  That would also give you the hydrogen/carbon information you need.  However, it would not detect gases like nitrogen.

The FID is a good all around detector that with the right column, might give you most of the information you need.  You might contact the instrument manufacturers application department for further inforomation.

Multiple detectors, columns and injectors can be put on the same instrument, if you need to mix methods.  This is a possible alternative.