-
- Posts: 1
- Joined: Thu Jun 24, 2010 8:12 pm
Hello, I've been in contact with Linda Koch of Supelco, and she recommended I contact Rodney George through this forum.
I am helping a graduate student with her analysis. She is conducting a reaction in a tubular flow reactor, mixing syngas (H2 & CO) with acetic acid dissolved in hexane. I don't know the ratio of gas to liquid, but could find out from her. With regard to the liquid only, she expects to see the following compounds in her product stream:
n-hexane (solvent) 98%
acetic acid ~2%
remaining compounds <1%
ethanoic acid
ethanol
acetone
acetaldehyde
methane
ethane
ethene
methanol
water
In addition to these, her product stream will contain unreacted H2 and CO, as well as CO2. We would like to be able to quantify all peaks (other than hexane) in the product stream. The reaction occurs at 300 °C and 450 psi.
We are using an Agilent 5890 GC which I set up previously for online analysis of Fischer-Tropsch synthesis products. There is a heated online sample tubing from the reactor, through a back pressure regulator, to the GC. The student is doing some calculations to see at what temperature she will experience condensation in her product stream, both for 450 psi and for pressure downstream of the regulator.
The online sample is analyzed in parallel on a capillary column (connected to FID) and packed column (connected to TCD), using two sampling valves. There is a third sampling valve installed, currently not in use, available for multidimensional work if need be. A pre-column connected to the packed column sampling valve holds up the heavier compounds in the sample, letting the lighter compounds onto the analytical column then backflushing the heavies.
It is important that we get peak areas for acetic acid that are representative of the concentration, since this area is used to calculate conversion for the reaction. I know that on nonpolar columns, the peak shape for acetic acid is not good (trapezoidal). Some chromatographers from other vendors recommended a PLOT Q capillary column. The peak shape would be somewhat better, and the PLOT Q could analyze the other compounds of interest. Linda Koch recommended an acid-treated column, like SPB-1000, to analyze the acetic acid. I suppose this would produce the most accurate concentrations for that compound.
If we go with the acid-treated capillary column, then ideally, hexane and the oxygenates in the product would be resolved and analyzed on this capillary column, and the other compounds (H2, CO, CH4, CO2, H2O, ethane and ethene) would be analyzed on the packed column. Hexane will be a large peak, and it is hoped the tail would not interfere with the earliest eluting oxygenate in the capillary chromatogram. For the packed column chromatography, we are currently using a 6' Restek Stabilwax column (packed with 10% Rtx Stabilwax) as a precolumn, and I'm curious if this will let water pass onto the analytical column. We are also using a 15' Carboxen column for the analytical packed column. The Carboxen appears to have issues with water, so am looking for a packed column that can handle water. For my FT work, I used two HayeSep D columns, one as the precolumn and the other as analytical column, and was able to quantify water.
What length and diameter of SPB-1000 column, for example, would be adequate for separating the alcohols, aldehydes and ketones listed? In the future, we would like to use this column to analyze cresol and guaicol as well, if possible. These would be substituted for acetic acid as substrates in the reaction.
Finally, I was looking at the NUKOL and SPB-1000 specs, and the temperature range is from 60 °C to 200/220 °C. Is it OK to operate the column below 60 °C? The packed and capillary columns are in the same oven, and the current temperature program begins at 30 °C.
Any thoughts on this analysis are welcome, thanks.
Kirk Snavely
Qing Pan
CEBC
University of Kansas
I am helping a graduate student with her analysis. She is conducting a reaction in a tubular flow reactor, mixing syngas (H2 & CO) with acetic acid dissolved in hexane. I don't know the ratio of gas to liquid, but could find out from her. With regard to the liquid only, she expects to see the following compounds in her product stream:
n-hexane (solvent) 98%
acetic acid ~2%
remaining compounds <1%
ethanoic acid
ethanol
acetone
acetaldehyde
methane
ethane
ethene
methanol
water
In addition to these, her product stream will contain unreacted H2 and CO, as well as CO2. We would like to be able to quantify all peaks (other than hexane) in the product stream. The reaction occurs at 300 °C and 450 psi.
We are using an Agilent 5890 GC which I set up previously for online analysis of Fischer-Tropsch synthesis products. There is a heated online sample tubing from the reactor, through a back pressure regulator, to the GC. The student is doing some calculations to see at what temperature she will experience condensation in her product stream, both for 450 psi and for pressure downstream of the regulator.
The online sample is analyzed in parallel on a capillary column (connected to FID) and packed column (connected to TCD), using two sampling valves. There is a third sampling valve installed, currently not in use, available for multidimensional work if need be. A pre-column connected to the packed column sampling valve holds up the heavier compounds in the sample, letting the lighter compounds onto the analytical column then backflushing the heavies.
It is important that we get peak areas for acetic acid that are representative of the concentration, since this area is used to calculate conversion for the reaction. I know that on nonpolar columns, the peak shape for acetic acid is not good (trapezoidal). Some chromatographers from other vendors recommended a PLOT Q capillary column. The peak shape would be somewhat better, and the PLOT Q could analyze the other compounds of interest. Linda Koch recommended an acid-treated column, like SPB-1000, to analyze the acetic acid. I suppose this would produce the most accurate concentrations for that compound.
If we go with the acid-treated capillary column, then ideally, hexane and the oxygenates in the product would be resolved and analyzed on this capillary column, and the other compounds (H2, CO, CH4, CO2, H2O, ethane and ethene) would be analyzed on the packed column. Hexane will be a large peak, and it is hoped the tail would not interfere with the earliest eluting oxygenate in the capillary chromatogram. For the packed column chromatography, we are currently using a 6' Restek Stabilwax column (packed with 10% Rtx Stabilwax) as a precolumn, and I'm curious if this will let water pass onto the analytical column. We are also using a 15' Carboxen column for the analytical packed column. The Carboxen appears to have issues with water, so am looking for a packed column that can handle water. For my FT work, I used two HayeSep D columns, one as the precolumn and the other as analytical column, and was able to quantify water.
What length and diameter of SPB-1000 column, for example, would be adequate for separating the alcohols, aldehydes and ketones listed? In the future, we would like to use this column to analyze cresol and guaicol as well, if possible. These would be substituted for acetic acid as substrates in the reaction.
Finally, I was looking at the NUKOL and SPB-1000 specs, and the temperature range is from 60 °C to 200/220 °C. Is it OK to operate the column below 60 °C? The packed and capillary columns are in the same oven, and the current temperature program begins at 30 °C.
Any thoughts on this analysis are welcome, thanks.
Kirk Snavely
Qing Pan
CEBC
University of Kansas
