base line wandering in temperature programming

[rajdeeproy]

Dear All

I am anaysisng some seawater samples for dissolved halocarbons by liquid extraction in pentane, but my base line in chromatogram is wandering and not stable. I am using a megabore comumn DB624(60m )coupled with HP 5 (5 m length) I am using varian GC with ECD but no autosampler. So i do manual injection
The Temp ramp program are as follwos.
40degree for 1 min
raised to 160 degree @4 degree per min
detector (ECD 220)
Column flow He 4ml/min
make up N2 (30 ml/min)

please suggest something

[Peter Apps]

Does the baseline "wander" only in one direction, or does it go up and down ?, how fast does it go up and down ?

Is the baseline wander reproducible from run to to run ?

Do you get a wandering baseline when you run a solvent blank ?, and is it the same wander as for samples ?

Does the baseline wander on a temperature programme blank, or between runs ?.

How (and why ?) do you connect the two columns ?

What is the purity of your carrier and make up gas ?. If you are using gas scrubbers, when was the last time that you changed them ?

Do you dry you rextracts before you inject them ?

Please post a chromatogram - instructions are on a sticky at the top of the LC postings.

Peter

[rajdeeproy]

Dear All

I am anaysisng some seawater samples for dissolved halocarbons by liquid extraction in pentane, but my base line in chromatogram is wandering and not stable. I am using a megabore comumn DB624(60m )coupled with HP 5 (5 m length) I am using varian GC with ECD but no autosampler. So i do manual injection
The Temp ramp program are as follwos.
40degree for 1 min
raised to 160 degree @4 degree per min
detector (ECD 220)
Column flow He 4ml/min
make up N2 (30 ml/min)

please suggest something

The base line goes up and down only in sample and not in standards and after 20min -30min the total run time is 42.

The base line wander is not reproducible form run to run .

The base line does not wander when i run a solvent blank or just a blank but the noise level incerase with the temperature ramp.

The two column are connected by fused silica connectors supplied by supelco and connected so that preseparation can take place in one .

The gas are of zero grade (99.9995%) which are further connected to moisture trap aor hydrocarbon trap.

The extract are dried by anhydrous sodium sulphate.

and Please tell me how to post a chromatogram so that it can be visually interpreted.

Can it be problem with the detector?

[Peter Apps]

To post a chromatogram go to the liquid chromatography forum, at the top is a sticky "Embedding chromatograms and reports", with what you have to do.

Or try clicking this link:

http://www.sepsci.com/chromforum/viewtopic.php?t=2617

Peter

[rajdeeproy]

http://i7.tinypic.com/25zqloi.jpg

[Bruce Hamilton]

There are clearly broad peaks from multiple earlier injections in there, you either should clean up the sample more ( SPE or a back wash/extraction ), or change the chromatography. Do they go away if the column is conditioned overnight?. Also try to inject the minimum amount necessary to see the peaks of interest.

You could try for a longer, higher ramp to drive them off the column perhaps with increased carrier gas head pressure ( my first choice ), add blanks between samples ( my second choice ), change the column so they come off quicker ( my third choice ) or install a backflush valve ( desperation ).

Bruce Hamilton

[rajdeeproy]

I have kept the column for condition at 200 degree and i get some peaks comming out. I will inject a sample after doing this and further report any changes. In the mean while i am also posting some more sample peaks please check these it might narrow down the possibilites.

thanks to all
http://i8.tinypic.com/261ccg0.jpg

http://i6.tinypic.com/261ccnt.jpg

http://i3.tinypic.com/261ccr7.jpg

[Peter Apps]

I would agree with Bruce, these a ghost peaks from earlier injections.

I addition to the fixes that Bruce suggested you could try using a much shorter column - you have lots of resolution to spare and so you could cut your present column into four pieces and still separate the peaks you are interested in (I presume that it is the four before 15 min that are the analytes).

Whatever you did on i6 looks pretty good to me !

The peak that elutes around 7 min has an odd shape though, and on i8 you have a negative peak at about the same retnetion time. There might be something that is quenching the detector.

Thanks for posting the chromatograms - it makes troubleshooting a whole lot easier !

Peter

[rajdeeproy]

The peak RT 12.06 and 14.06 are of interest. The first one is chloroform and second one is carbontetrachloride. The broad peak which appears at 7 in two image comes only if i use sodium sulphate to dry the sample extract.
For my standards i was using methanol. Should i use dry methanol or acetone? which one one will be better.

The quenching of the detector which peter mentioned is seen only if i dont use anhydrous sodium sulphate for drying?

Can i be sure that my detector is ok and not contaminated?

Now the column total length is 70 m (both the column included) I will cut the DB624 coulm (15m each and will try to use)

I will post the developments

Thanks for all those suggestion

[Bruce Hamilton]

Thanks for posting the chromatograms, very helpful.

Because your detector is only seeing some of the compounds ( if you split the column outlet and send to FID as well, you will see the organics, but not any water ), quenching may occur because of coeluting peaks.

Your standards should be processing the same as your samples, and in the same solvents. Linearity should be good for both peaks of interest, and you don't have to worry about other peaks, besides ensuring they elute in the same run .

You need to ensure that reducing the resolution will not cause coelution. The detector should not become contaminated - provided it is at a reasonable temperature ( usually recommended by instrument manufacturers ).

I'd usually select a thin film column with low bleed for ECD of a few halogenated molecules and, in my experience 624 columns tend to bleed more than many other phases. Low bleed = low background, less problems and detector contamination, and usually means you can inject less.

I'm not sure where you method came from, but if it works please keep using it.

Bruce Hamilton

[rajdeeproy]

Dear Peter and Bruce .

Thankyou very much for all those useful advice. The base line has stop wandering and i am getting good peaks. I will post few of those. I am also giving the details of the method iam following.

Determination of biogenic and anthropogenic volatile

halocarbons in sea water by liquid—liquid extraction and

capillary gas chromatography

Katarina Abrahamsson and Silke Klick

Department of Analytical and Marine Chemistry, Chalmers

University of Technology and University of Göteborg, S-412

96 Göteborg Sweden

Received 20 December 1989; revised 28 March 1990. Available

online 4 January 2002.

Well i have modified this method to suite my GC's technical specification but the basics are from here.

My name is rajdeep and i work in National Institute of Oceanography Goa as an research fellow and doing my Phd in phytoplankton pigments and biogenic emmision of halocarbons and this is my second year in research. So i feel i am very young and still a long way to go.

best regards
rajdeep

[rajdeeproy]

http://i2.tinypic.com/2639cb9.jpg

http://i7.tinypic.com/2639cgg.jpg

[rajdeeproy]

http://i2.tinypic.com/2639cb9.jpg

http://i7.tinypic.com/2639cgg.jpg

[Peter Apps]

Hi Rajdeep

That is a dramatic improvement ! What did you change ?

Peter