Negative peak overlapping CH4 peak

[aki]

Hi everyone!

We use a Varian-CP 3800 (Column type Porapak QS, 80-100 MESH, Dimensions: 4mx1/8"x2.0mm SS; split valve at the end of the column that separates flow between ECD and FID). Nitrogen is used as carrier gas (nitrogen generator). FID settings: hydrogen 30ml/min, nitrogen 30ml/min, air 300ml/min, detector 300ºC, oven 30ºC, injector 70ºC.
Problem: The FID is showing an additional peak exactly overlapping the CH4 peak (in standards and air samples). This additional peak is separated into a strong negative and a less strong positive part; furthermore the peak is drifting out at the end so that the end of the peak area is hard to define. The negative part of the additional peak is getting less strong over time but is not disappearing. Repeated analysis of standard gases shows high standard deviations (standard deviation for 5 samples between 0.21 and 1.37).
Changes that were made so far: Change in parameters like injection volume, injection speed, column temperature, injection temperature, air flow and flow rate. The column and detectors were baked for cleaning. The flame tip was checked. None of these changes did help.

Any ideas?

Many thanks!

[Spuzzin]

Check your valve timings. Negative peaks on gas analysers are often caused by valves kicking in. From a development point of view getting the valves to activate at the right point (away from target compounds) is usually the trickiest part.

Rich

[aki]

Thanks!

There is no valve timing I am aware of. The valve is only a splitter which does not change during the run and merely halves the flow between the FID and ECD.

[larkl]

Don't think there's much that you'd see by FID this early in the run. O2/N2 would elute, but of course the FID doesn't respond to these. Really does sound like a valve issue. Could it be an injection phenomena, pressure spike?

[AICMM]

aki,

Larkl's point about pressure spike is very valid, especially if doing a hand injection using a syringe. Nothing on the ECD at the same retention time?

Best regards.

[aki]

Hi,
we observe a strong positive peak at the beginning of the run time on the ECD but is is before the retention time of CH4 - the ECD does not show anything at the same retention time. We do not inject manually but use a combipal autosampler.
Thanks
Aki

[Peter Apps]

An FID is remarkably robust to different operating conditions, but I think that there might be a gas flow rate problem. You have your FID make up gas (nitrogen) set at 30 ml/min, which is the default optimum. In addition to that you have more nitrogen (how much ?) flowing through the column, and then when you inject you have a pressure increase in the inlet depending on the injection speed that will cause a flow pulse through the column. This might all add up to a deviation in the FID baseline that looks like a tailing peak. Try reducing the FID makeup flow so that the total nitrogen flow is 30 ml/min, and reduce the injection speed to equal the carrier gas flow rate.

Peter

[aldehyde]

Negative peaks on ECD are often indicative of contamination (contaminants absorbing ions/electrons before they make it to the electrometer). Not sure about negative peaks on FID though.

[jcgc]

Hi there,

http://www.flickr.com/photos/21889311@N ... 475173085/

We still have this problem, anyone have any new insight?

Thanks

[AICMM]

jcgc,

How much are you shooting and what concentration of methane? You have two negative peaks in the FID chromatogram which is unusual without some kind of external event like a valve event, although the first might be the air peak?

Along Peter's line of reasoning, are you adding 30 mL of make-up or is that your column flow rate and, if column flow, are you adding make-up?

Best regards.

[jcgc]

Hi,

750ul, I had a column flow of 43ml/min and make up of 30ml/min so I dropped the make-up to zero. I have noticed an improvement in the negative peak if I run the fid for a few hours. I have to turn if off every night as we are using a H generator and bottled O, does an FID normally take hours to stabilize and if the gas lines to the GC are turned off each night, could this lead to any diffusion contamination?

[AICMM]

jcgc,

Do you leave the FID hot overnight? If so, it really should not take all that long to settle in (usually.) Also, are you using air or oxygen, by your post? How low an injection volume have you tried? Trying to syringe inject 750 uL can be a problem based on what I have seen with one of my customers although they were using micro-packed, not 1/8". This is especially true if you try to do it fast. That customer went to a valve injection. Finally, please clarify the splitting between detectors, a "T" right, not an actuated valve?

Best regards.

[Don_Hilton]

I would make an injection of pure nitrogen - from a source other than the generator. See if you get the negative peak. If so, I would consider contamination of the carrier gas and displacement by a band of non-responding gas in the sample. And the reason I would take the pure nitrogen from a source other than the nitrogen generator is that the nitrogen stream would be the suspect source.

While a pressure spike would still be a possibility, discssion of this changing over time would leave me suspecting the gas quality.

By the way: How clean is the air going into the nitrogen generator and how long since filters were changed?

[simsa]

I also have a negative peak in my FID before the CH4 peak. This peak is O2. Maybe you can get rid of it by decreasing the carriergas flow or increasing the lenght of your column to ge a better separation of the two peaks