I'm just started using a GC, don't have much knowledge about it. We are using this old old HP5890 Series II. But we changed almost everything. New liner, column, rings, washers, you name it. We are trying to detect derivitized amino acids.
The conditions I'm using are :
Temp 110-320 @32deg/min
column length 10m
diam. .25 (not sure of the unit)
uses a normal FID
can't think of anything else I know about it.
Well, the problem....when I inject, I get this HUGE peak, which is so broad it washes all the other peaks out. It starts at about .4 seconds and ends near about 4 mins.
So, I kept the purge on the entire run duration. I still get a massively high peak at about .4 seconds, but it's narrow. Sometimes the peak splits into two or three all about .4 to .9 seconds.
Any Ideas? ANY help would be GREATLY appreciated.
I guess I could just continue here about the next problem. I prepared alanine solutions of different concentrations (0.5 mM, 1 mM, and 2 mM). We have this packaged derivitization process, which we add 6 reagents (No idea what they are, the wonders of patents) and shake. So, I do the derivitization process, and then run the three samples.
Once there was a peak at about 1.3 s, which is around the expected time of alanine. This exact peak repeated on a subsiquent pure Ala-Ala run, confusing me even more. But now, consistantly, those three samples do not give large peaks. I get a response of about 8000 for the 0.5 mM and about 2000 for the 2 mM. Why does the response go down for increased concentration? It doesn't make any sense to me. I hope I have made sense to someone out there.
A very confused soul!
By Russ on Monday, July 22, 2002 - 01:35 pm:
To run a split injection, the purge must be on during the run. If you have the purge off, you are doing splitless and until you turn it on you will get a very broad solvent peak. By the way, leave the purge on when not in use. Leaving the purge off when not in use will likely cause premature failure of the solenoid. Have you measured your linear velocity or column flow rate? Generally with a 0.25 mm ID column the linear velocity is measured instead of flow rate. Does the GC have the electronic pressure control or does it use pressure valves? What carrier are you using? Is it the same as the one specified in the method you are using?
By Rajiv Luthra on Monday, July 22, 2002 - 08:45 pm:
Okay, So I'll keep the purge on all time it. My column flow rate is about 32cm/s. (That sounds okay right?). Yes, I have electronic pressure control, and it's set up to use constant flow as required. My carrier is Helium, and that's what they say to use.
By Anonymous on Tuesday, July 23, 2002 - 01:34 pm:
Too many unknowns. You don't know if the GC works correctly, if your chemistry is working, what the retention times of known standards are, and I'm sure you mean "minutes" instead of "seconds" for the peak retention times you mentioned.
Make sure the GC works right by running another material that you can simply weigh or dilute, and NOT derivatize. See any of the GC catalogs for example chromatograms and compounds for your column. Only then can you start working with unknown materials. Can you get genuine derivatized standards?
You must take the problems one at a time, and the GC must work correctly, because that is the only way you can detect the chemicals.
By Rajiv Luthra on Wednesday, July 24, 2002 - 11:01 am:
Thanks for the advice! Yes, 1.3 minutes, not seconds! The internal standard I'm using has a supposed retention time of about 2 mins. The column and derivitization procedure are a commercial package from Phenomenex. They aren't much help. The column can't be used for anything else, or at least that's what it says. We ran the GC with chloroform, and it seems to respond fine. Double the volume, double the peak height, and the results are reproducable.
I'm looking into getting some genuine derivitized sample. That seems to be the best idea.
Any other ideas?
By Anonymous on Wednesday, July 24, 2002 - 12:30 pm:
Yes. Hopefully there is someone in your lab who could verify that the 5890 is working properly. You might be able to dig up a Grob test sample from someone in your lab (or pg. 37 of the Phenomenex '99/'00 catalog), or throw together something similar from the chemical shelves. Sometimes test mixes with chromatograms and conditions are shipped with new columns. Perhaps Phenomenex will send you a known good sample (if it's stable) from their applications lab, or perhaps they have a saleperson or chemist that could take a look at your system with you.
Chloroform is sometimes used as a solvent for other analytes in GC work. Observing a linear response for it doesn't prove much to me, since it is present at 1000's of times the concentration of analytes in your samples.
Old 5890's are pretty simple machines. If the pressures and temperatures are right, and you get a signal when you inject, I would expect it to work (but no idea about response for any given material--we use about 0.5 to 2 ugs/injection of pesticides on our FID's). I have no idea about your particular application, but it sounds like Phenomenex is trying to make it turn-key. You sound like a good test-case! If I worked for them, I'd sure be curious about what kind of troubles you might have with the product.
I'm also bothered by the fact that this doesn't seem to be a mainstream analysis, as evidenced by the lack of example chromatograms in the catalogs of any major supplier (that I have). Maybe it isn't as simple as it appears. Chemistry can often be like that!
Posting is currently disabled in this topic. Contact your discussion moderator for more information.