Page 1 of 1

Practical Limit in FID?

Posted: Sun Aug 07, 2016 12:31 pm
by mattmullaney
Good Morning,

In HPLC-UV, the "signal out" is sometimes regarded as having a practical maximum of 1.0 AU. I understand that GC-FID is based on a rather different physical phenomenon--regarding the detection. Along that line, though, is there a practical maximum detection output in FID that should not be exceeded?

Or is the only consideration to ensure that it is possible to appropriately model the upper concentration end of a given calibration while employing FID?

My thanks for your help, in advance.

Re: Practical Limit in FID?

Posted: Mon Aug 08, 2016 7:30 pm
by rb6banjo
Isn't that "1.0 AU" cutoff because of the way absorbance is calculated based on the %Transmittance?

A = log10(100/B)

so at B = 100 (all incident light is getting through to the detector), A = 0. When B = 10 (only 10% of the incident light getting through), A = 1.0.

The logarithm of this ratio goes extremely nonlinear once you get larger than A = 1.0.

Image

The red line was the y = mx + b fit for calculated absorbance of 1.0 or less. It's mostly linear when A is less than 1.0.

Analytes that respond in a flame detector are flammable. The detector will have a nonlinear region as well and it will be determined by the ability of the instrument electronics to differentiate between the ion current in the flame for the analyte as a function of concentration. For the FID, it is very wide (thankfully). I haven't thought about it much before now. I'm usually operating at the opposite end of the linear dynamic range. I'd bet it'll be analyte dependent as well.

Re: Practical Limit in FID?

Posted: Mon Aug 08, 2016 9:12 pm
by mattmullaney
Hey rjbanjo,

A succinct and thorough explanation which I thank you for and admire quite a bit.

Take Care!

Re: Practical Limit in FID?

Posted: Tue Aug 09, 2016 12:11 pm
by lmh
I think you may be doing modern UV/Vis detectors a slight disservice! They are quite capable of calculating logs and showing the true absorbance as calculated from an initial transmitted intensity. They don't (necessarily) lose linearity at high absorbance merely because the transmitted signal is varying logarithmically.

On the other hand, the logarithmic thing means that every doubling of OD (doubling of concentration) makes the signal 10 times harder to detect. It also means that the gradient of signal-change per concentration-change is ten times less, so yes, the sensitivity of the instrument to changing concentration will fall. Of course, on top of this, if their response isn't actually exactly logarithmic at low intensity, and the electronics/software are based on perfect logarithmic obedience to the Lambert-Beer law, then there will be loss of linearity.

Basically, the 1.0 limit depends on the quality and dynamic range of the actual detecting thing (photodiode array, or whatever), and the quality of the amplifier that handles the initial tiny signal. As technology has improved, these have improved, and there are instruments out there now where the old 1.0 rule is a bit pessimistic.

Re: Practical Limit in FID?

Posted: Tue Aug 09, 2016 12:29 pm
by itspip
I was under the impression, from a former coworker/boss now retired, that the 1.0 rule came about due to the analog nature of the detector. That the max voltage output of the detector was 1V which was assigned to 1.0 AU and that the modern detectors have a much larger linear range.

As far as the FID upper limit, you could always make a large range concentration curve. I think most of us are down at the other end and are more concerned about how to make it more sensitive.

Re: Practical Limit in FID?

Posted: Tue Aug 09, 2016 1:34 pm
by rb6banjo
Absorbance is calculated as it's calculated. The detector records how many photons are hitting it when there's nothing in the beam and then it records how many photons are hitting it when there's something in the way. The difference determines the percentage of light transmitted. I'm not sure that fact has changed since the "old days".

A = log10(100/B) [when B = 100, A = 0]

The graph I created is a plot of B on the x axis and A on the y axis. This function is really only linear when B is in the 10-100% range.

If the detectors are still photon counters, I don't see how you can overcome this mathematical limitation without some type of artificial compensation in the software. Perhaps that's what instrument manufacturers do to increase the linear dynamic range? I don't know. My UV-VIS is about 25 years old. I generated this data a while back in conjunction with a project.

https://1drv.ms/b/s!AkH-uI0tnY5LddwtZJcO-6kGfAk

This detector is quite linear for the range of concentrations of the dye that gives a response no larger than A = 1.0 (middle of the visible region of the spectrum, pink line on second page of the pdf). You could probably extend it to A = 1.2 or 1.4 without suffering too much. It starts to go nonlinear after that. It's still predictable but it is decidedly nonlinear.

Re: Practical Limit in FID?

Posted: Tue Aug 09, 2016 1:46 pm
by mattmullaney
I stand properly chastened (nothing like forums such as these to promote humility in mattmullaney!) on the score of 1.0 AU.

Folks, my thanks and appreciation. The FID isn't based on a physical phenomenon that relies on something akin to optical absorbance behavior...perhaps "practical limit" isn't what I really mean to say and I mean "theoretical limit" to FID. I think rjbanjo has what I'm looking for in his answer and agree 100% with you last two fellows as well.

What you also have noted is proper--even in the use of UV-Vis detection, we still have to go and do the experiments to see what happens at the high and low ends of our proposed range of quantification.

Re: Practical Limit in FID?

Posted: Tue Aug 09, 2016 3:11 pm
by GOM
Hi

Interesting question and replies

With regard to your query about the FID detection level

is there a practical maximum detection output in FID that should not be exceeded?

did you really mean lower rather than upper limit?

a) upper limit is not an issue - it is more governed by the capacity and resolving power of your column for capillary columns to get symmetrical peaks from the amount injected

edit - to bring your sample within this range would typically require dilution of the sample and /or splitting of the injected amount

b) Linear range is in the order of 10 to the 6 to 10 to the 7 orders of magnitude

c) minimum detection level generally is in the order of nanograms to picograms injected depending on the molecule and background noise. The FID is a mass sensitive detector. That means that its response is proportional to the mass of carbon that passes through it. In that regard, FID response is stated in terms of picograms carbon per sec. Detection limits for FIDs are in the low pg C /sec.

Regards

Ralph

Re: Practical Limit in FID?

Posted: Tue Aug 09, 2016 5:46 pm
by tom jupille
Basically, the 1.0 limit depends on the quality and dynamic range of the actual detecting thing (photodiode array, or whatever), and the quality of the amplifier that handles the initial tiny signal.
The other factor is stray light. If you have 1% stray light hitting the photodetector, it will *never* calculate an absorbance greater than 2.0 (and will go non-linear well before that).

Yes, a modern detector should be built in such a way that minimzes stray light, but after it's been in use for a while, and the flow cell has been removed and cleaned a few times, and the gaskets on the monochromator may be getting tired, . . . , staying below 1.0 AU is sound advice.

Re: Practical Limit in FID?

Posted: Wed Aug 10, 2016 1:21 am
by mattmullaney
Hi Ralph, Tom and everyone else responding,

My thanks for your consideration to you all. I greatly admire the clarity of comments with respect to the UV-Vis detector, particularly with regard to the "extension" of monotonic response behavior at absorbance values greater than 1.0 AU. It is my failing that could not describe this adequately in the recent past to other folks that asked me to do so.

@ Ralph, nope...I was asking about a maximum value of FID response. I can see how what I was asking of is more than a little outside of the norm, though--I'm also concerned about the "low end" of the calibration curve, just as "everyone" else is. And yes, this is a second question asked of me that I was less than able to explain to others that asked me to do so.

Hard for me to imagine, that after 20+ years, that I have:

1. Still much to learn.

2. What I think I "know" I am unable to explain properly. Kind of makes me feel a bit of despair.

Reminds me of teaching labs in grad school...was answering these questions, thought I was doing pretty well...and was deflated completely when not only could I not make a point clear to a student, but when another TA came in and had his turn at explaining the same point, the student UNDERSTOOD the concept.

I think that the adage "If you cannot simply explain a thing, you may not know it well enough to begin with" may be applicable in my case...but then, it does give me something to continue to strive for.

Pleasant Evenings, folks!