new in the developing method - volatiles GC-MS

[dana_]

Hy,
I am new in the field of chromatography, so any suggestions would be helpful. I am developing a method for volatiles in cigarette smoke. I have done calibration curve. My quality control is perfect, and my correlation factor in calibration curve is < 0,995 for all compounds.
My sample is a referent cigarette for which I have results from other laboratories. Average results from 20 laboratories are 1,3-butadiene 60,3 ug/cig, isoprene 553 ug/cig, acrylonitrile 12,3, benzene 60,3 ug/cig and toluene 85,4 ug/cig. My average results from many analysis are 1,3-butadiene 75 ug/cig, isoprene 600 ug/cig, acrylonitrile 13 ug/cig, benzene 45 ug/cig and toluene 40 ug/cig. Peaks in the chromatogram are looking good. They are not tailing and the baseline is not raising. Do you know why I do not have the same results like other laboratories? What can be the source of this problem?
I am using a column DB-5MS 60m x 0,25 mm x 1 um. Do you know how long is the chemical good for analysis when it is opened (like toluene 99,5% or benzene 99,0%)?
Thank you!

[Don_Hilton]

First question - did you smoke the cigarettes under standard conditions. This includes room humidity, proper humidity in cigarette storage, standard puff cycle, and proper linear velocity of airflow across the cigarettes while smoking? All of these have a big effect on smoke chemistry.

[dana_]

Hy,
Thank you for reply.
Yes, I did all like it was written in method. I also tried smoking with 1,2 and 3 impingers, and I smoked this referent cigarette many times, and I get the same results. I tried to change the liner, because I do not know what is the influence of different type of liners (all for PTV inlet) to my analysis. I hope I will get some results today, even if it is not liner, so I can exclude that.
If you have any more idea, please let me know.

[Peter Apps]

Your results show discrimination against the two heaviest compounds. This is most likely to be due to inlet discrimination. What are your analytical conditions ?

Peter

[dana_]

I have PTV inlet (liner: single baffle with glass wool, deactivated).
Analitical conditions:
injection volume: 1 uL
split 20:1
injection port temperature: 250 °C
column flow: 1.0 mL/min
column temperature: 35°C for 10 min
20°C per minute to 260°C, hold 10 min
transfer line: 250°C

Column: first column: DB-WAX 1 m x 0,25 mm x 0,25 um
second column: DB-5MS 60 m x 0,25 mm x 1,0 um

I tried with a different liner, without glass wool, and with multibaffle liner, but I got worse results. So I do not really know what I can change in the inlet.

Thank you
Đana

[Peter Apps]

How much glass wool is in the liner ? Some of them are stuffed with way too much. You need just enough to help the solvent evaporate evenly - say about a 5 mm length of the liner loosely packed with wool just below where the needle tip reaches to.

What solvent are you using ?

How are you calibrating, and how do you make up your calibration standards ?

Peter

[Don_Hilton]

What method are you using? Is it a CORESTA method? One published elsewhere?
And, what method was used for the values you are trying to match? Any differences in equipment or consumables?

[dana_]

Peter,
I am using a liner with about 10 mm glass wool. It is Agilent liner, and I can not change the glass wool because it has ID 1,8 mm. I am using methanol as my solvent.
Benzene d6 as internal standard

Primary standards:
Analit Target weight (mg) target volume (ul) weight (mg) Purity Volume (ml) Final conc. (ug/ml)
Isoprene 1000,0 1500,0 992 0,99 50 19641,6
Acrylonitrile 150,0 190,0 152 0,995 50 3024,8
Benzene 250,0 280,0 266 0,999 50 5314,7
Toluene 150,0 185,0 162 0,999 50 3236,8
Styrene 100,0 110,0 87 0,999 50 1738,3


Secondary standards:

Analit Aliquot of primary stand. (ml) Final volume(ml) Conc. (ug/ml)
Isoprene 5,0 982,1
Acrylonitrile 5,0 151,2
Benzene 5,0 100 238,5
Toluene 5,0 159,2
Styrene 5,0 102,0

5,0 ml of primary stock (each compound) - dilute to volume (100 ml) with methanol.
I separatly made primary and secondary stock of 1,3 - butadiene.

I have 7 standards. Using aliquot of secondary stock, and 1,3-butadiene (0,05 ; 0,10; 0,5; 1,0; 3,0; 5,0; 7,0 mL) and 250 uL internal standard in volume of 25 mL.

Don_Hilton:
I am using Arista because I did not have a column which is written in Coresta. I am still waiting for it, so when it arrive I will try use Coresta. The only difference is in column. The results I am trying to get was all manage by method of Coresta.

Thank you all

[Peter Apps]

These analytes are very volatile, what precautions do you take to ensure that your standards have the composition that you calculate ? Are you working be weight or by volume, how do you dispense them etc.

Injecting methanol into a hot inlet causes a serious pressure pulse, and this is probably causing an unstable split ratio that is the cause of the discrimination. You could try diluting your solution x 20 and then injecting splitless - you will need a wide bore liner to accommodate the methanol vapour. A larger liner might even help in split mode.

Peter

[dana_]

I weight my flask with methanol before and after I put a compound in it (volume).
Can you tell me, if I should order a different liner, like PTV liner, high temperature (quartz or borosilicate) ID 3,4 mm, volume 713 uL or 668 uL. I am now using liner that it is not noticed that is for high temperature, and my inlet temperature is all the time 250 °C. I will try to dilute my solution x 20, but for now I only got this liner (PTV liner, single baffle, deactivated, 2 mm, volume 200 uL).
Thank you for this idea...

[Peter Apps]

250 C is not a high temperature in inlet terms - above 350 C would count as high. You do not need quartz.

Peter

[dana_]

OK, thank you so much.

[Don_Hilton]

Does your GC have a split/splitless inlet on it? This would allow you to use a liner with greater volume. And, does the Arista method tell the inlet type used on the GC? I'ts been a while since I did tobacco work, so I don't remember the level of detail in an Arista method - or if I even had one of their methods pass through my hands.

Doing a quick look, I noticed the following: http://www.beitraege-bti.de/pdfs/2011-24-04-174.pdf. And, while I Don't have time to look it over thoroughly I'll pass it on to you. You have probably already seen it. But if not it may be helpful.

The article suggests spiking the traps and workign up the samples as though they came from cigarettes. This may help determin if the issue in your method is in smoking or in workup and analysis of the material in the traps.

[dana_]

Thank you, I have seen it before. My new column has arrived (which is written in Coresta), so I will try now to develop a method by Coresta.
I will post the results when I finish.
For now thank you all for your help.

[dana_]

Don_Hilton:
I have a split/splitless inlet on my instrument. In Arista it is not written which inlet is used. But I need a PTV inlet to set my temperature in oven, right? Or I can do it even with split/splitless inlet?
Thank you