How to start with HSGC?

Chromatography Forum: GC Archives: How to start with HSGC?
Top of pagePrevious messageNext messageBottom of pageLink to this message  By Anonymous on Monday, May 3, 2004 - 02:52 am:

Hello,

I want to use a 7694 HSS with a HP 6980 GC and a 5973 MSD to replace a GC(FID)-methode with a headspace/MS method.

My samples contains about 10 different VOC in a not reproducable liquid matrix. Another problem is, that i have more and less volatile compounds like 1-Propanol and n-methylpyrrolidone.

I have some problems to determine the "right" amount of sample to put into the vials.
I tried 0.25g of sample + 0.04g of an internal standard in a 22ml vial.

Then I did a sequence with incrementing equalibration time [5min; 45min], but still can't determine the correct time at 80C vial temp.

Q:
Is it possible to develop a HS-methode for a mix of several VOC (more and less volatile) in a difficult matrix (paint)?

Is it a good idea to increase vial temp, when the less volatile compounds have to less area/response?


Top of pagePrevious messageNext messageBottom of pageLink to this message  By Anonymous on Monday, May 3, 2004 - 06:15 am:

A complete evaporization method (using a minimum of sample) would probably work best for you.

I would also use a much smaller vial if possible.

Large vials and samples do not give you a better RSD or Limit of detection.

Review Anal. Chem 1997, 69,2221-2223 for small sample HS analysis, ie 25L or less.

You will need a higher temperature to measure the n-methylpyrrolidone well, in my opinion, perhaps, about 100-120C.


Top of pagePrevious messageNext messageBottom of pageLink to this message  By Anonymous on Monday, May 3, 2004 - 07:32 am:

Thanks, I'll try that tomorrow.
But what about the amount of internal standard. Should I work with an ISTD that appears "between" the less and more volatiles.
What amount of ISTD makes sence, when I use 25l of sample?

"Review Anal. Chem 1997, 69,2221-2223"

Is this a HSGC method? Where can I get it, nothing found with goolge! Can i download it?

Thanks a lot


Top of pagePrevious messageNext messageBottom of pageLink to this message  By Anonymous on Monday, May 3, 2004 - 10:47 am:

An internal std is generally not necessary if your system is operating properly. However, to prove your system is working properly, any volatile analyte can be added as long as it mixes with your sample well and is volatile at the temperatures you wish to use.

I would dilute paint in a suitable solvent with an internal standard like n-butanol.

The article is a GS-MS method using a 1mL (no split used) fixed loop HS analyzer Tekmar 7000 and 6mL or 2mL vials and 0.53mm capillary columns in a Varian 3500 GC with deactivated loop and transfer lines. Vials were pressurized (to 6 or 10 psi?) Carrier flow was 20cc/min with a 3 meter 0.25mm ID FSOT detector side retention gap. A slower carrier rate can be used (5 or 10cc/min).

Perkin Elmer does carry the adapters to use std 2mL autosampler vials in a 22mL vial autosampler. See someone in that company for details.

I don't have an electronic file of the article available at work. I do remember that the use of smaller vials allowed a great reduction in LOD and improved RSD of analytes that were present at less than 1ng per 25mL of sample. Linearities were > 0.9995 for 18 solvents with recoveries of > 95% when measured at a 50ng spike of analytes. Most recoveries were 98% or better. Pyridine was the worst I believe.

Methanol, ethanol, acetone, acetonitrile, ethyl ether, methylene chloride, ethyl acetate, isopropanol, n-propanol, pentane, hexane, chloroform, benzene, sec-butanol, MEK, n-butanol, pyridine, toluene, trichloroethylene, and dioxane were some of the validated analytes used with this method, which used either water or dimethyl acetamide as a carrier solvent.

Range of measurement was 1ng to 1000ng of analyte per sample. Heating at 80C was for 6 minutes with internal shaker. Oven temperatures were starting at 40C and then programmed. Cycle time was 15 min with cool down but toluene eluted in less than 6 minutes. The unidentified early peak in the example chromatogram was later found to be acetaldehyde.

Be careful measuring esters and aromatics with butyl rubber HS cap liners. Use aluminium lined (prefered) or at least teflon lined caps. Don't use aluminium lined caps with chlorinated analytes or carrier solvents.

Sorry I cannot remember more details. It has been 8 years since the research was completed.

Good luck


Top of pagePrevious messageNext messageBottom of pageLink to this message  By ced on Monday, May 3, 2004 - 11:17 am:

HI


What are your VOCs?

I would also dilute 1/10 the paint to reduce the matrix effect,in a suitable solvent..dmac;dmf?

After you should try in a 10 mL HS vial to put 1 ml of your sample diluted.
And repeat with 5 mL sample,compare the area to see if there is an increase,this is not always the case,depending on partition coefficient from the VOC.

Increasing the temperature,help to reach faster the equilibrium in your vial,what is less important for more volatile,but necessary for less one such n-methyl pyrollidone,then you need to extract sufficient substance from the liquid phase to detect it.

Depending on your solvent i would try 30 min 80C and 1h 80c and compare the aera increase.

See You


Top of pagePrevious messageNext messageBottom of pageLink to this message  By Anonymous on Tuesday, May 4, 2004 - 12:50 am:

There are about 10-15 VOCs (1-Propanol, 1-Butanol, 2-Butanol, 2-butoxyethanol, diacetonethanol, ethandiol, n-methylpyrrolidon, ethylhexanol...)

I'll try the complete evaporization methode today. But I haven't read much about it. Why is this methode not used everytimes with HS, when it reduces the problems with more sample amount?

But I also try to dilute the paint. I'll post my results, then.


Top of pagePrevious messageNext messageBottom of pageLink to this message  By ced on Tuesday, May 4, 2004 - 01:00 am:

The main purpose of FET ,full evaporation technique,is to eliminate the matrix effect,but the remaining residue can caus e adsoprtion effect!
Regular HS-GC has a higher sensitivity..

See You

More info in
"STATIC HEADSPACE-GAS CHROMATOGRAPHY ,Theory and Practice"
B.KOLB,L.S.ETTRE
Wiley-vch


Top of pagePrevious messageNext messageBottom of pageLink to this message  By Anonymous on Tuesday, May 4, 2004 - 01:46 pm:

I would try using water to dilute the paint if water soluble or a hydrocarbon [pentane/dimethylbutane/hexane/cyclohexane] (depending upon the VOCs you are trying to measure) if not water soluble. If DMF or DMAc separate from your VOCs then use one of them.

Sample a small portion (25-100L) in a 6mL or smaller vial. I would heat the vial for a maximum of 10-12 minutes at a temperature of 85C.

A teflon lined cap should work nicely.

You should have a rather low limit of detection, a lot less than you probably need.

Anonymous from Monday, May 3, 2004 - 10:47 am:


Top of pagePrevious messageNext messageBottom of pageLink to this message  By Anonymous on Wednesday, May 5, 2004 - 06:50 am:

I noticed a type in my first response. The referred method is a GC-HS method not a GC-MS method. Sorry.

Anonymous from Monday, May 3, 2004 - 10:47 am:


Top of pagePrevious messageNext messageBottom of pageLink to this message  By Anonymous on Thursday, May 6, 2004 - 02:58 am:

Thanks, complete evaporation works pretty good with my standards.
My vial temp is 100C, loop 110C and transfer line 120. Equalibration time is 10min. Sample amount is 10l. Seperation seems pretty good, too.

I also did a sequence with a dilution of 1g(STD)/25ml, but the result wasn't ok with some compounds. I'll try 5g/25ml in PEG400.


Top of pagePrevious messageNext messageBottom of pageLink to this message  By Anonymous on Thursday, May 6, 2004 - 07:59 am:

We all enjoy good news. Thanks for telling us the end of the story.

Anonymous from Monday, May 3, 2004 - 10:47 am:


Top of pagePrevious messageNext messageBottom of pageLink to this message  By Anonymous on Monday, May 17, 2004 - 07:00 am:

Hi,

I'm still working on the HS.
I got some problems with my calibration. For most of the compounds (1-Prop., 1- and 2-butanol) i got a calibration curve that isn't linear.
If i choose linear regression, coef. of det.(r) is about 0,95. When quadratic regression is chosen, the curve fits great through all my standards with a r of 0,99x.
I tried ten standards [1;10], but if i have to analyse all of the standards every time, I can forget this method.

Q: Is this usual with HS?
I don't know how it depends with direct injection with out HSS. Also non-linear?
Q: What causes this non-linearity?

Quantification works with quadratic regression but not with linear regression forced through (0,0).

Ethandiol is the only compound, that's more difficult to seperate with a good width and is very near to my ISTD (MIBK).


Top of pagePrevious messageNext messageBottom of pageLink to this message  By Anonymous on Wednesday, June 9, 2004 - 10:04 pm:

hi

I am working on a very fundamental method development project for Ethanol analysis. I am using HP 5890 series II with FID. I am unable to get a linear curve with a high regression for my method. for my method, i use varying ethanol mass fractions , diluted to ratio of 1:10 and then shot in GC. I am not using any internal standard.
Please provide information if any for developing a calibration curve such that high ethanol concentrations can be analyzed.
Thanks


Add a Message


This is a private posting area. A valid username and password combination is required to post messages to this discussion.
Username:  
Password: