GC With FPD-detector: no response.

[rikkert]

Hello all,

I have a GC and recently installed a FPD-detector on it. We are now trying some samples on the detector (sulfur containing samples) but we can't get peaks with 'low' concentrations.

I made a test sample with 10% thiophene (in cyclohexane) and the detector works fine, it only gives one peak for the thiophene.

But when I analyse a monster of sulfur containing diesel oil I don't get any peaks. The total sulfur (measured with XRF) is 1500 ppm. This concentration is spread over several compounds (I don't know what compounds or how many) so it makes sense that the compounds aren't 1500 ppm individually but maybe there are 10 compounds with 150 ppm each.

How is it possible that I don't get any peaks while the detector should be able to detect trace levels of sulfur.

My setup:
GC: Chrompack CP9001
Injector: Splitless
Column: 25M CP-sil 5 cb (0,32 mm ID)
Detector: FPD detector model 906a
Temperature program:
-4 min Initial phase (70 degrees celcius)
-30 degrees/min Rise
-60 minutes Final phase (280 degrees celcius)

[Peter Apps]

The most likely cause of the problem are the parameters that you do not provide us with - the gas flows and settings for the new detector.

Peter

[rikkert]

The gases are set as required:
- Air 1: 40 ml/min
- Air 2: 165 ml/min
- H2: 142 ml/min

The detector settings:
- There is NO baseline subtraction.
- Range is set to auto range.
- Output attenuation is set to 16.
- Recorder offset is set to 10 (which corresponds with 10 mV)

I think that's all. Correct me if I'm forgetting anything;)

[AICMM]

rikkert,

Sounds to me like you may have a dual FPD?

Split or splitless would be really good to know. If split, split ratio would be good to know.

FPD's roll off in response very quickly so while you may see the 10% the response at low ppm will be substantially less. This is because of the nature of FPD response (quadratic response.)

FPD is also subject to quench so 150 ppm component in diesel may be quenched by the diesel.

Finally, drop your attenuation as a first step. You are significantly decreasing the signal to recorder with the attenuation.

Best regards,

AICMM

[rikkert]

Hello AICMM,

I don't know what you mean by 'dual FPD', we have an FPD with 2 flames (if that's what you mean).

The injector is split/splitless (so we can set it to split but in this case we didn't use the split function) so the injection was Splitless.

With quadratic response you mean to say that when a sample has a 4x higher concentration, the response will be 4*4 times higher?

I have tried to lower the attenuation (it was much higher at first) but I will try to lower it even more (maybe even NO attenuation).

I don't hope the Sulfur compounds will me quenched by the other compounds, that would make the detector useless for our purposes.

[GasMan]

The problem with the FPD is that the sulfur signal can be quenched if the sulfur compound co-elutes with a hydrocarbon. This is very likely if you are trying to find sulfur in diesel. The trick with sulfur analysis is get the sulfur compounds not to co-elute with hydrocarbons. This will mean carefull choice of column and possibly multi column switching.

Gasman

[AICMM]

rikkert,

Considering you are splitless, you are putting a lot of diesel on column. Therefore, quench is very possible. Yes, double the concentration can mean close to 4 times the response which makes low level detection difficult, especially if not resolved from hydrocarbon. Dual FPD's are supposed to be a little less susceptible to the quench but I have not actually done that so.....

You might try two dimensional GC since sulfurs could be more polar (and held up more) in the second dimension. Have not seen anyone publish this but it's a thought. Do you need speciation or just total sulfur? If just total, then look at instruments like Antek or Elementar for total sulfur analysis. Used all the time in petroleum industry. If you need speciation then you might have to look at something like the SCD.

Best regards,

AICMM

[CE Instruments]

rikkert,


You might try two dimensional GC since sulfurs could be more polar (and held up more) in the second dimension. Have not seen anyone publish this but it's a thought. Do you need speciation or just total sulfur? If just total, then look at instruments like Antek or Elementar for total sulfur analysis. Used all the time in petroleum industry. If you need speciation then you might have to look at something like the SCD.

Best regards,

AICMM

He has already done total Sufur by XRF
As he is in Europe he could consider Total S from the manufacturers of what was the Thermo TS/TN http://www.tshbenelux.nl/

Running an old CP9001 will have quite an old FPD and probably not the most sensitive. If you can see your standard then try diluting it gradually down to the levels you need to see. Then if successful try a 50/50 diesel standard mix, do you get quench from the hydrocarbons?

2D application note from Restek for Sulfur in Diesel with TOF MS detection http://www.restek.com/pdfs/PIN11_JULY-GCXGC.pdf
I think you would need to bin the Chrompak GC first

More modern FPDs and pFDPs get to similar levels as SCD, perhaps a new GC may be all that is necessary

[rikkert]

Hey guys,

Thanks for all the replies.

After some experimenting with the diesel on the GC (and some different temperature programs) I have succes with measuring the sulfur.

Thanks for the help, and till next time

[Peter Apps]

Hey guys,

Thanks for all the replies.

After some experimenting with the diesel on the GC (and some different temperature programs) I have succes with measuring the sulfur.

Thanks for the help, and till next time

So how about feeding your new knowledge back to the forum ??

Peter

[rikkert]

Hello Peter Apps,

Sorry for the late response, I just saw you comment and noticed that it would be nice to share my experiences with the forum.

The reason why we wouldn't get any response with the FPD is that we were using such a long temperature program that the components came out as really broad peaks. But the 'peaks' were so broad that it looked as if it was just the baseline.

With a much shorter temperature program we got only one peak at first (one really wide peak which contained all the sulfur components), then we fine-tuned the temperature program in such a way that we got a nice separation and sharp peaks.

Also the we lowered the attenuation of the detector. Because we use an analog>digital converter to retrieve the detector signal on a computer we get noise from the converter. Small components get quenched in the noise but by lowering the attenuation from the detector all the small components get 'bigger' but the noise from the converter remains the same.

That wraps it up pretty much. Hope this helps anyone in the future;).