above the baseline GC-FID Chromatogram peak

[kankyo]

I am analyzing various petroleum products using GC-FID. I wonder why hydrocarbon peaks move above the baseline as you increase the temperature. I can see peaks that touches the baseline at 40C for 5 minutes hold. As I increase the temperature at a rate of 10C/min up to 260C, peaks begin to be elevated and move above the baseline. Why does it happen? I understand that for a single or few peaks you can easily adjust the baseline. How about for samples like hydrocarbons with numerous peaks do we have to adjust the baseline from the base of one curve to another? Or we just take as it is.

[CE Instruments]

First run your column with no sample just to see where your baseline actually is, it may well not be flat. Secondly do you actually get baseline resolution of your peaks ? In Hydrocarbon mixes there can be more peaks than your column can resolve and it apears that your baseline rises however it is just that the peaks co-elute. Most data systems do a good job integrating, you do not mention the data system you use ? Can you post a chromatogram for us to comment ?

[kankyo]

Thank you CE instruments.
May i know how to post a chromatogram here. Baseline without the sample and with the samples are the same. It is just that the hydrocarbon peaks rises in the middle of the chromatogram. Is this co-elution? In computing for the concentration of the hydrocarbon peak, shall we include the area below the peaks( the area caused by co-elution?)? Or adjust the baseline to the base of the peaks.

Again, a million thanks.

[kankyo]

When run as a single hydrocarbon it touches the baseline. As you hold the temperature from the very beginning the peaks touches the baseline. As the oven temperature increases, it begins to rise. Is co-elution of peaks only happens in hydrocarbons with many peaks? what causes co-elution. I am using EZChrom Wlite Version 3.1.5

[pueblito]

Dear Kankyo.
You could download the EPA 8015 method:
http://www.epa.gov/epaoswer/hazwaste/te ... /8015c.pdf.
On page 33 it is especified what kind of integration you have to use, It is called baseline-baseline integration.
Well, It is very difficult for a single column to resolve all of components in a sample of a petroleum product , because of that we see a "hump" ( for diesel) with large shape peaks of n-alkanes that coelute with branched and other type of hidrocarbons in a defined range of boiling points.
before calibrating be sure the baseline is estable.

bye.

[CE Instruments]

To post a picture or chromatogram you first ned to have it in digital form. THe easiest method is to use a photo hosting site like photobox. Import your digital picture or chromtagram scan/ data output (jpeg) Underneath you will see several boxes that are links to the picture. Copy the picture link including the [img] options and paste into here.

If your baseline without peaks is the same as the baseline with peaks then you are resolving the peaks and you should quantify from the baseline. If the trace without peaks is flat i.e. no rise then the peaks should be quantified to an artificail line drawn by the data system that corresponds to the "blank" chromatogram[/img]

[kankyo]

Thanks for replies.
I posted three images of chromatograms to the following link:
http://www.photobox.co.uk/my/album?album_id=53652857.
Massive co-Elution in chromatogram 1 begins at 7 minutes and ends 18 minutes. There are still small peaks that rise above the baseline from 18 to 21, are they still hydrocarbon peaks or just noises?
At lower hydrocarbon concentration as in the case of Chromatogram 3, many peaks rise above the baseline so that when you compute the area you will get a big value.

Thank you for further comments

[Bruce Hamilton]

Firstly, your samples contain several hundred HCs, and those major peaks are presumably n-alkanes. The samples look like either an unrefined gasolines, kerosine or diesel fractions.

You are very unlikely to obtain baseline resolution of the compounds, and it almost looks as though your integrator settings are not suitable. You have to also decide what you are trying to achieve.

If you are looking for boiling range data, you don't require that level of resolution, and there are ASTM methods available that define the columns and conditions for measuring boing range of different fractions and crudes. They often just use grouped peak areas within timeslices that have been calibrated with standard mixtures. Virtually all column suplliers have data on simdist columns and methods on their www sites.

If you are looking to characterise the HCs, then you need to establish the true baseline of your column when solvent is injected and the overn temperature ramp rate to give maximum separation for your samples is being followed.

Setting the ramp rate is a bit of an art, depending on film thickness, sample boiling range and composition ( which define the initial and final temepratures , but most column manufacturers offer some initial guidelines.

To start, for medium film and 50 metre long columns, suitable resolution for gasoline fractions occurs at about 3 - 5C/min ramp, and 5 - 8C/min for kerosine and diesel, starting about 30-50C below the initial boiling point. Usually initial hold periods are relatievly short, just a few minutes maximum.

You need to ensure that you perform multiple blank injections after conditioning the column to establish the baseline for the temperature progamme you are going to use.

You then inject a standard mixture to help set the basline - most GC suppliers, ( and the ASTM ), can identify suppliers of HC standard mixtures if you don't have one, or can't make one.

You may have to force the sample baseline to follow the profile of blank, using various baseline and peak sensitivity integration parameters.

If the blank baseline increases during or after the sample sequence, you need to revisit the oven programme, especially the final temperature and final hold time.

There is plenty of information in the column suppliers literature, and some suppliers ( esp Restek, Agilent, Supelco ) have a lot of information on petroleum anaysis available for free on their sites, and also give good support to clients with their columns.

Please keep having fun,

Bruce Hamilton

[kankyo]

Thanks a million to all your ideas, suggestions and comments.

[Ron]

On the first two chromatograms the tallest peaks are straight chain hydrocarbons, the smaller peaks are aromatic hydrocarbons, and possibly thiophenes, and the rest is what is sometimes referred to as the unresolved organic component, but is usually just called "the hump". The third chromatogram has some aromatic peaks, but is primarily just the hump. These look like fractions of crude oil, but the third one seems to be highly degraded. These aren't bad chromatograms for this type of sample, the resolution actually looks pretty good.