Poor linearity FID Shimadzu GC2010

[EUROFINS]

Hello Jake,

I am also glad to hear I am not the only one with this problem. I spend a lot of time figuring out what might be the problem and I am convinced something is wrong with this detector.

Best regards

[Jake]

I can't offer a solution, but it is very surprising that changing the makeup gas makes a difference to the FID's linearity. Any idea why ?

I'd love to say that I have a great explanation, but I don't. Switching to He make-up gas was the very last thing we considered and changed as we made our system set-up more like the application note here:

http://www.shimadzu.eu/products/chromat ... diesel.pdf

Again, the quality of the data improved moving from N2 make-up to He, but the problem persists. As one would expect, there was no difference moving from H2 to He carrier. But I guess that's what one would expect for switching make-up gases, so we thought it prudent to confirm.

@EUROFINS - Good luck with your Shimadzu rep. As I complained here about the problem (I'm in the US), I was hoping the issue would make its way to Japan. I eventually gave up.

And just to clarify our situation: If you plot the area response versus the amount of a given FAME, or even the total area versus the total amount of FAME, the correlation of determination (r2) is > 0.999. But the problem is with the slope - double the amount of FAME does not give twice the signal.

Keep us posted.

[EUROFINS]

Sorry, mistake.

[Jake]

Yep. There it is. Nicely demonstrates what we're experiencing too.

[EUROFINS]

Hello Jake,

My Dutch Shimadzu representative would like to talk to you and your Shimadzu representative about your experience with the GC-2010 FID. Could you contact Wouter van Dijk on **31736430320 or visit the Shimadzu website; http://www.shimadzu.be/contact/

Thanks

[EUROFINS]

The same detector is used in the GC-2014 (for capilary columns) so the GC2014 probably has the same problem.

[Jake]

Hi Jan-Hendrik,

Sorry for the delayed response. I did get the email from Tom about this thread being updated, but I was having problems logging in.

I did fill out the info at the Shimadzu.be link. A bit tough for me to call since I'm on the west coast of the U.S.

I'm really curious to hear what you've found out. If you want to keep it out of a public forum, please send me a PM.

Our GC-2010 has been preforming beautifully, aside from the detector issue. It's running 24 hours/day, at least 6 days/week, and has run 10,000+ samples with just your typical routine maintenance.

John

[EUROFINS]

Still working on it. I am thinking about spreading the news under the 150 EUROFINS labs worldwide.

[aldehyde]

When I first read your posts my thought was I would focus more on the inlet and ALS/syringe than the detector, but in this case you could be right to suspect something with the detector.

I have heard before that adjusting the flow ratio/increasing flow in an FID can extend linearity for higher analyte concentrations. I just googled quickly and found this: http://www.chromtech.net.au/pdf/AdvHints-FID.pdf

I think this is where I originally read it, so take it with a grain of salt. But you can give it a try.

Sensitivity can be increased by decreasing the hydrogen flow
rate relative to the carrier and make-up gas. Usually, a 1: 1.25
ratio of hydrogen to carrier gas (25c&nin H,, 30cc/min carrier +
make-up) increases sensitivity. However, at lower hydrogen
flow rates, the flame is prone to extinguishing as large volumes
(>2pl) of water elute. A higher than optimum hydrogen flow
rate is usually necessary when using water as a solvent to ensure
the flame remains lit at all times. Increasing the air flow rate
above 300cc/min. has little effect on FID sensitivity. However,
increasing the air flow rate to 400-450 ml/mm. will extend the
linearity of the FID when high concentrations of analytes are
injected. Since there are variations in detector design, consult
your instrument manual for optimum gas settings.

I have also heard that for some applications N2 as make up can increase sensitivity, but I don't have a citation for that.

[AICMM]

Eurofins,

If I understand you correctly then, you did various split ratios to determine linearity? Is this the case?

Regarding nitrogen versus helium make-up, I have found one subtle reference to that improving the sensitivity and I have experienced it myself although it is less dramatic than some people made it out to be. (I will try to find the reference in my office when I return.)

Regarding changing gas flows to change the linearity, my experience has been that it certainly makes a difference changing the flows at very high carbon concentrations (between 0.5% and 100% methane concentrations for example.) That is not 4 orders of magnitude (by any shake) but changing the flows made the signal in that range much more consistent.

Best regards,

AICMM

[DSP007]

Such was also observed in our Agilent 6890. While working version the next. If hexane obtained by distillation of crude oil - it contains an admixture of higher molecular hydrocarbons (a mixture of isomers).
If we dissolute test solution in hexane - then concentration in the solution of the relative share of impurities in the peaks and the proportion of noise in the peak area increases. This worsens the scattering of the points and among individual determinations, and leads to the exit from the linearity.

[EUROFINS]

I found, using nitrogen as a make up gas, sensitivity increased almost 100% compared to helium.
I varied spit ratios, injector temp, type of liner, injection type etc. but it had no influence on linearity. The only parameters that influenced linearity are detector gas flows and jet diameter. I widened the hole of the 0.3 mm jet to 1 mm and observed a dramatic improvemend of linearity but sensitivity decreased aprox. 50 %. Because of the mutch better linearity we are now running the samples with this 1 mm jet. Shimadzu has a 0.8 mm jet availible for water based analysis.
It looks like some other brands suffer the same poor linearity.

Thank you very mutch,
Jan-Hendrik Tolsma

[Peter Apps]

Hi Jan-Hendrik

Whan you say that sensitivity with nitrogen increased to almost double that with helium, did you measure signal:noise, or only the peak size for a given quantity of analyte ?

One of the models for the operation of the FID is that highly active hydrogen radicals from the flame diffuse against the gas flow (i.e. towards the jet) and exhaustively fragment and hydrogenate analyte molecules coming through the jet. It is then the products of this hydogenolysis that are combusted and ionised in the flame proper.

This model could make sense of the difference between helium and nitrogen as make-up, and of the effect of jet diameter, if gas flow velocities and diffusivities had a significant effect on upstream diffusion. I have no idea whether the gas velocities at a flame tip could have this effect.

Peter

[EUROFINS]

Hello Peter,

Sensitivity is expressed as peak hight or area here. Noise increased about 15 %.
I certainly belief diffusion has something to do with it. I thought in the direction of the ionisation current measurement. Gas flows influence the distance between the flame and electrodes: jet and collector.

Does your theory explain the decrease in signal when mounting a 1mm jet?

Jan-Hendrik

[Peter Apps]

Hi Jan-Hendrik

The theory about the diffusion of hydrogen radicals is not mine, unfortunately I left my literature on FIDs at my previous job so I cannot give the reference.

With a 1 mm jet the gas flow velocity is slower (by nearly an order of magnitude) than with a 0.3 mm jet. I suspect that the shape of the jet tip, and whether it induces turbulence in the gas might also be relevant.

If I recall correctly, acceleration of the ions in the flame under the influence of the polarising voltage is much faster than diffusion, and so would not be influenced by gas flow.

Peter