Chromatography Forum

Hey, I'm a student working on a lab report and for a GC experiment. For some reason I can't seem to understand why my graph looks weird.

I uploaded my raw data if anyone wants to check it out. I'm still trying to fix my printer so I can scan the results from the chromatogram and upload those as well.

http://www.mediafire.com/?dgrei4oxup7qzla

Hi Shinghan

The question that I am about to ask is not a rhetorical one, and it is not intended to embarass you. You answer will help the discussion under the thread "Quality of the forum".

So the question is: How can we tell you what might have gone wrong if you do not tell us what you did ?

Thanks Peter

Tell us please:

Did you use an autosampler or did you inject and start the runs manually?

Did you use helium carrier or nitrogen? or some other gas?

Describe the ID and length of the capillary column you used.

Your flow rates seem to be in error by two orders of magnitude, 7µL/min at 22cm/sec linear flow rate?

How did you measure the peak widths and the retention times?

As you might have guessed there may be an error in your work. Help us help you to find it.

Thanks,

Rod

Oops! Haha sorry about that. I've never really done anything like this before.

The instrument used was a Varian GC.

The column is a Restek RTX-5 30m x 0.25 mm x 0.25 µm.

Helium carrier gas was used for the analysis.

The samples were manually injected by hot needle injection.

The a few of the retention times were labelled on the chromatogram, but for the majority of the samples, the retention times were calculated from the graph by hand. The same was done for the peak widths, calculated by hand.

I hope that's enough information, but I doubt it since the calculations for the peak widths I have done are on the chromatograms. Later today Ill head to the library and scan them.

EDIT:

Didn't make it to the library, but I found a way to get the pictures of the chromatograms.

The test mixture that was used for the analysis contained n-C13, n-C14 and n-C15. The compounds were dissolved in carbon disulphide solvent. The results below refer to the isothermal runs with increasing linear velocity of 12, 20, 30, 50 and 70 cm/s.

http://img825.imageshack.us/i/isotherma ... arvel.jpg/
http://img600.imageshack.us/i/isotherma ... arvel.jpg/
http://img827.imageshack.us/i/isotherma ... arvel.jpg/
http://img571.imageshack.us/i/isotherma ... arvel.jpg/
http://img716.imageshack.us/i/isotherma ... arvel.jpg/

These images are closeups of the n-C14 peak which was needed to plot a Van Deemter curve.

http://img9.imageshack.us/i/isothermalr ... eupof.jpg/
http://img812.imageshack.us/i/isotherma ... eupof.jpg/
http://img101.imageshack.us/i/isotherma ... eupof.jpg/
http://img841.imageshack.us/i/isotherma ... eupof.jpg/
http://img46.imageshack.us/i/isothermal ... eupof.jpg/

I just want to make sure I've got the right idea in mind to create the plot.

So first, the air peak time is calculated using the Peterson and Hirsch method. The retention times can be found directly from the chromatogram which is the time from injection to elution.

Second, the peak width is measured at either the baseline or at the half-height of the peak (I used both in excel to validate the answer) using a ruler. I used ratios here to calculate the peak width (For example, if the scale was 2.2 cm = 0.1 min and the width I measured is 3.4 cm, then I would put it as a ratio and solve getting 0.1545 minutes).

The number of theoretical plates can then be found by using the peak width and retention time.

Next, the plate height is found by using the column height (30m = 30000mm) and the number of theoretical plates.

I was asked to plot height vs. the actual linear flow rate for the C14 peak, so I would find the average linear velocity first? Then change it to linear flow rate?

EDIT 2:

Ok so I changed up my units to make them more simpler, but my hopes to that accomplishment would enable me to find the reason why my graph looks quite different from most graphs. I played with the values to see how it would affect my graph and it turns out that if I change the theoretical plate height for sample 4, the graph resembles much like a Van Deemter plot. I was suspecting that the problem was hidden with either sample 4 or 5 and it turns out that sample 4 is a possibility. I'm now trying to understand how one could get to this kind of conclusion. For the plate height to be affected that way, the number of theoretical plates has to change and so does the retention time and the peak width. Right now Im looking into what could affect the retention time and peak width.

Heres my edited data:

http://www.mediafire.com/?q0zuehghfi8i9x8

Hi Shinghan

Since you are a student, you have to do the thinking. Have a look at the chromatograms and see if you can spot anything strange about one of them - hint; the vertical axis is your friend.

Is there a connection between the odd chromatogram and the deviant data ?

Kudos for doing manual injections and for actually measuring peaks rather than blindly relying on autosamplers and software readouts.

Peter