help on supelco application note 164

[mmegliol]

Hello everybody,

I'm trying to set up the method on the application note on my gc, i'm using the same column but with helium.

[url]
http://www.sigmaaldrich.com/Graphics/Su ... 0/8661.pdf

Using butyl acetate as a INTSTD for all the component what could be a ggod ammount to add?

I try to calibrate the method adding 20 and 50 microlit, tha calibration curve are preatty good (r2 is 0.999 0.998) for all the component but is i try to reinject the standar i'm getting different ammount from the standard.

I'm injecting 1microlt with split ste up at 30.

Thanks.

[Peter Apps]

Manual or automatic injections ?

Peter

[Don_Hilton]

A few questions:

Are the early eluting peaks and late elluting peaks behaving the same way as far as irreproducability? What linear velocity are you using with the helium? How is the peak shape on your peaks?

On how much internal standard: Ideally, the internal standard peak size should be about the size of peaks you are trying to quantitate. Make it small enough that it does not tail or front from overloading and large enough that it give good and reproducable areas.

[mmegliol]

I'm performing a automatic injection with the autosampler, all the peak are wel sepaerated with no tailing at all, but i think the way I add the internal standard is not OK!

I took from a chromatography book the informatation to add the internal standard with approximately the same area of middle point of calibration.

When I perform my calibration I add 20 microlt to all my 4 standard level, at the end of the calibration i took a look to the area and between the 3 injection of methanol (for exemple) on the same level I have 2 in std%, but if I look averall to all the injection of the butyl acetate (internal standard) should be allways the same but between the injection of all the calibration level i have 25 in std%.

I think the error is on the addtion of the 20microlt, probably I need to prepare a higher conctration of stock solution and add more solution to my 4 level calibration in order to reduce the error on the volume, is this correct???!?!?

thanks a lot!

Matteo.

[Peter Apps]

A 25 % error in adding 20 ul is high. How are you doing it ?, syringe, pipette, autopipette ?

If you dilute the internal standard in order to add a larger volume to the samples or calibration standards you will have to make a correction for the dilution of the sample by the added standard.

Peter

[mmegliol]

Hi petr thanks for your help!

usaully when i prepare the standard for the injection i make a stock solution with methanol butanol and all the other component I'm looking for, i dilute in 4 flasker with different concetation and 10% ethanol and I add 20 microliter of butylacetate with the pipet, after that i bring the solution to volume at 100ml.
I also try to ioncrease the volume of the internal standard to 50 micolt, and change pipet but did not improve.

I was thinking to prepare a stock solution of the internal standard at 881 ppm and take 5 ml to add at every level of the calibration, this way the concetration of the internal standard won't change but i'm using a bigger volume? is this a good idea?

thanks.

[Don_Hilton]

If you are using a pipette that has the disposble plastic tip and it uses an air column to lift the solution - or butyl acetate: These are good for aqueous solutions, but bad for use with volatile solvents. The vapor pressure of a volatile solvent will push the solvent out of the pipette.

Use a glass syringe with a steel needle. For 40 or 50 microliters, use a 100 microliter syringe. If you do not have one of those, weigh the butyl acetate into a volumetric flask and add solvent the line. (Or add solvent to give a total weight.)

[Peter Apps]

Don is right - use a syringe not a pipette.

Peter

[mmegliol]

thanks a lot guys I'll try to use a different syringe!

Do you think would be better useing a glass syringe even to prepare the alchol solution (methanol, propanol atc...)?

thanks.

[Peter Apps]

For volumes up to 500 ml a micro-syringe is best, or for the smaller volumes disposable glass microcapillaries, for volumes 0.5 ml and above a glass pipette is OK (but can be tricky to use for very volatile liquids), for 5 ml and above you can use a glass pipette for additions and volumetric flasks for final volumes. Autopipettes with disposable tips that have a plunger inside the body and air above what you are measuring are VERY difficult to use accuratley with volatile liquids.

Peter

[mmegliol]

I try to use gas sirynge to dose the internal standard and component and i also change the sirynge on the autoinjector from 10 to 5 microlt.

right now i'm injecting .5 microlt abd i'm having %std on the area of my componet included the internal standard that avarge around 10, a bit to high for the autoinjector.

Is it possible having lower std% considering that I'm injecting a water base solution (ethanol10%)?

thanks.

[Don_Hilton]

I am not entirely sure of your question, but I'll try.

First thought: Be sure your inlet maintantanace is up to date. A leaking septum adds all kinds of variabilty that you don't want.

Now that we have addressed the easy one to fix, on to the harder ones.

If your question is variability of the area of the internal standard and it shows an relative standard deviation of 10%: I would look at integration conditions to be sure the peak is integrated the same way every time. I would also try using a 1 uL injection rather than 0.5 uL. If everything is optimized as good as you can get it, 10% RSD on the internal standard area may be as good as you will get. I would try things like a slightly hotter inlet -- but with water, you quickly run into the risk of your expansion volume putting analyte out the septum purge vent or back into the inlet plumbing. A hot inlet with a pressure pulse could work as well - but that takes some tinkering and may not get you any benefit.

If your area ratio of analyte to internal standard shows a 10% standard deviation, I would look at the same issues. I like 5% RSD's on methods. (But if I am doing trace analytical, I may be happy with higher RSD's - to get some kind of result at ppb levels.) Run five injections of the same standard - mid range in your curve - one after the other and let me know how the areas vary for internal standards and a couple of analytes - one early and one late. I may ask you to post chromatograms.

Unfortunately, this is one of those problems I solve best by sitting in front of the instrument, and examining all of the parameters and talking to the instrument (It does not talk back - but I feel better).

Don

[mmegliol]

thanks everybody for the suggestion I'm almost done with the method, I had to increase the temperature at the injection and decease the volume injected to achive good std% right now i'm at 1.5, I think is a good resoult.

Last question!!!!!

I'm almsot ready to perform the calibration but during the last series of injcetion i notice the The methanol comes out before the acetealdehyde but in all the chromatogram on the application note is reverse, is it possible that higher injction tmperature or using helium in styead of nitrogen will switch the 2 peak?

thanks again to everybody!

Matteo.

[Don_Hilton]

A change of carrier gas itself, probably not. If you have a dramatic change in flow rate, maybe. Having noiced that columns have changed over the years (even though labeled the same) another possibility is the column may be slightly different that the one used in creating the apps note.

Any change in injection order resulting from a change in inlet temperature would have to be the result of partitioning in the inlet - which I would expect to lead to broadened peaks. And the peaks in the apps note would not look very good.

You say that you have reduced volume injected and are getting good RSD's. Which kind of syringe and what volume did you settle on? And -- an RSD of 1.5% is great. Good job!

[mmegliol]

Hi Don,

thanks for helping again!

I start using a 5 microlt syringe in stead of 10. Using the 10 and injecting 0.5 was not reproducible at all!!

Where can i send you a cromathogram to take a look, just to know what you thinking!

thanks.