Any tricks to increase linear range of Agilent MSD?

[Anthony_Ng]

Hi all,

Recently I read a technical note about the determination of Phthalates:
http://www.chem.agilent.com/en-US/Searc ... &liid=2633
(Agilent Note: 5990-4863EN)

In the note they can produce a wide linear range from 0.25 to 10ppm. However I can only got a very narrow range, from 0.25 to 0.75ppm. Is there anything wrong for our MSD? Is there any tricks I can do to make a wider linear range?

By the way, do you think 10ppm is too much for MSD? Ususlly I get a flat-top peak when the concentration is around 4ppm. Don't know how do they do that in splitness mode.

[Ron]

What is your electron multiplier setting? If you are getting flat topped peaks it sounds like the multiplier is set much too high. If you are setting a fixed value lower it by 200V and see what the result is. If you are setting relative to tuning result, which is what I prefer, try running with an offset of 100V or less.

[Anthony_Ng]

The EMVolt, after autotune, is around 1600V now. I found that the method setting adds 400V on top of the tune result. (i.e. around 2000V now). My colleagues said this method is copied from a pesticide analysis method.

Is it too high? So that the linear range to become too narrow also?

Thanks in advance.

[carl.nott]

Is this a 5975? How low can you go? Are you getting '0.25 tiny peak, 0.50 med peak, 0.75 huge peak'?

[tima]

How does this translate to nanograms on the column??? The concentration levels that you quote must be factored for sample weight, final extract volume, injection volume and then you get your ppm.

So a 5973 should get a much wider dynamic range. Firstly, get someone to check over your dilutions or calculations because a 5873 should be able to reach as low as 50 picograms and as high as 1 to 2 hundred nanograms range (this is the "on-column" injected with no sample factors applied to the detection). The range is usually good.

I think you maybe injecting too much. How is the peak shape?? If it rised slowly then tails quicker. Are you overloading your GC column? What kind of column are you using?? Lots of questions.

Timothy

[Anthony_Ng]

Thanks all. Let me give you more detail information:

Our model is 5973N, not the newer 5975. And what we prepare is ppm in a sample vial. (i.e. 0.25-0.75ppm in a vial, and inject 1ul of it by autosampler).

Column: DB5ms, 30m x 0.25id x 0.25 (The most common one)

Yes, I can see 0.25 tiny peak, 0.5 med and 0.75 big and even up to 3 or 4 ppm will be a very large peak. The problem is the linear range is narrow (e.g. from 0.25 to 0.75ppm, or from 1 to 3ppm. It looks curve upward when we plot from 0.25 to 3ppm level.)

The peak shape is fine, even for 3ppm, not much fronting and tailing. Therefore, as tima said, does the application note I quoted is not the "concentration in the sample vial"? Is it the "concentration of phthalate in the sample"?

[Don_Hilton]

A curve opening upward often indicates losses of material in the inlet or column. Be sure you have a clean, deactivated liner. (New out of the box!) If you are using glass wool in the liner - it should be will installed before deactivation. And be sure the column is in good condition. If active sites are part of your problem, they will limit your dynamic range on the low end.

[aysal]

Hello,
for most of the solvents, determination of phthalate is really difficult. Contamination of solvents by phthalates effects whole process, even std calibration. If your solvent (used for std prep or extraction) contains phthalates, you have already have some response in std blank. Do you have any signal for your std blank? If it is so, what is the diffrence in response between the blank and first calibration std?
You may also use + 200 emv, then you get at least 1-2 more std in your calibration table?
Application note mentioned about starting point of calibration for 0.25 ppm and 0.5 ppm. If it is OK, then you just check your 0.25 ppm std for S/N ratio. Do you use same ions as note mentioned for calibration (307 for DIDP)? What is your R2 value for 0.25 to 0.75 ppm and 0.5 to 3 ppm? May be you have low response in 0.25 to 0.75 range that can not fit upper level of calibration? or too much response then you should go lower concentrations.
For extraction method you have more concentrated samples comparing standarts? İf you do not have more dilution, you will have only 25 fold dilution of samples, i.e. if your samples contain 1000 ppm phthalates, after extraction you get 40 ppm in vial.

Ayhan

[Ron]

There is something really odd going on here, the 0.75 ppm peak should be 3 times the area and approximately 3 times the height, and that does not seem to be the case from the qualitative descriptions above. What is the threshold setting in your method? The things you are seeing would make a lot more sense if you have a very high threshold setting and are clipping off the bottoms of the peaks.

[Anthony_Ng]

Thanks all. The following was what I have tried yesterday:

I made some injections by varying the EM Volt (+0V, +200V and +400V on top of the autotune value, 1600V). The 3 calibration curves still not liner in these 3 cases. All of them curve upwards. (concentration: 0.1, 0.5, 1, 2, 3ppm in sample vials. Injection vol: 1ul, splittless)

Ron, thanks for your suggestion. Let me check it when I back to the lab tomorrow. What is the most usual/general setting of the threshold?

Is the linear range of phthalates very narrow actully? If so, our MSD may not have problem......

By the way, I just wondering how the application note can inject 10ppm phthalates in GC/MSD without getting a flat-top peak?

[Ron]

1 uL of 10 ppm will not give a flat topped peak unless the mutiplier voltage is set too high.

[gpronger]

There are a few things we are fighting when we're trying to do low-level calibrations and maintain a reasonably wide linear range. If we simply inject less we can run into inlet problems causing analyte loss (which would lead to the type of curve you are experiencing). Injecting more can "swamp" the active sites in the inlet, but then we lose our linear range.

With the 5973N (and I'm assuming a 6890 GC), there are a few injection options you can explore to mitigate some of these issues.

First, a pulsed splitless injection will decrease the time the analytes are in the inlet so that you'll see less degradation. I would try this option first. I would try around 30PSI for 0.5 min. This should give a decent transfer. An Agilent paper on the technique can be found at:

http://www.chem.agilent.com/Library/app ... 9944EN.pdf


A second option is to take advantage of the gas saver valve, but to allow you to operate the inlet in a low split to high split mode. This will allow you to inject more (2uL), not overload the column, and then sweep the inlet for any remaining residue. To do this, I would set the split flow to around 1:3. Have the gas saver "Off" as typical. At 0.5 minutes, turn the gas saver "On" but have the flow go to 50mLs/min. What this allows you to do is increase the injection volume to minimize active sites in the inlet, while not overloading the column. We tend to run our 5973N in this mode for traditional EPA 8270 and 625 type analyses.