Validation Issue - VOC's in gc/ms Agilent 7890A/5975C

[Bigbear]

Are you using purge and trap?
Please post your GC conditions as well as P&T ( if using) trap temps. flow rates.

[Steve Reimer]

Surrogates and internal standards are required for the method although the compounds used can be chosen to meet data quality objectives.
Link to 8260C: http://www.epa.gov/epawaste/hazard/test ... /8260c.pdf

To fix the vinyl chloride non-linearity you need to find out if the problem is chromatographic, with your standard prep or something else. For that we need more information.

[TiagoBele]

Thank's for the reply bb and steve,

My extraction method it's static headspace, heating at 80 degrees (celsius) for 15 minutes.

The system it's automated (CTC-Combpal), with a injection volume of 500 uL of the headspace, with the split mode of 10:1. The method i'm using it's the SIM, i had ran a SCAN just for sure of the compounds, the retention times, and to verify the qualifier ions for the SIM mode.

After that my method it's very similar to this one :https://www.chem.agilent.com/Library/chromatograms/442A.jpg


the column is the same.

The differences it's that my gas flow rate is 0.6mL/min, and it's not the purge and trap.

Ramp :

Initial : 44 C for 6 min,

4C/min - 170 final temperature.

Total : 38.5 minutes.


Steve . "Surrogates and internal standards are required for the method although the compounds used can be chosen to meet data quality objectives." - so, i can validate my method without using surrogates and internal standards, but maintaining a quality control with the compounds standards that i'm using? is that it?


Thank's to all for the reply, Peter , BB and Steve.

[Peter Apps]

That Agilent method uses 5 ml of purge from a 100 ng/ml standard - you are trying to get down to 1 ng/ml with 0.5 ml of headspace. Lets suppose for the sake of argument that you have 20 ml of headspace in your vial, and all of the volatiles are in the headspace. If you have a 1 ml sample that is 1 ng in 20 ml of headspace. You take out 0.2 ml, containing 0.025 ng of analyte, and inject that to the GC with a 10:1 split, putting 2.5 pg on the column. r-squared of 0.98 down to those levels is remarkable, and should encourage people to buy equilibrium headspacers rather than P&Ts.

Peter

[TiagoBele]

That Agilent method uses 5 ml of purge from a 100 ng/ml standard - you are trying to get down to 1 ng/ml with 0.5 ml of headspace. Lets suppose for the sake of argument that you have 20 ml of headspace in your vial, and all of the volatiles are in the headspace. If you have a 1 ml sample that is 1 ng in 20 ml of headspace. You take out 0.2 ml, containing 0.025 ng of analyte, and inject that to the GC with a 10:1 split, putting 2.5 pg on the column. r-squared of 0.98 down to those levels is remarkable, and should encourage people to buy equilibrium headspacers rather than P&Ts.

Peter

Hello Peter!

I don't know if you guys in south africa have knowledge in brasilian portuguese language, but here is an article published in a respected journal here in our country:

http://www.scielo.br/scielo.php?pid=S01 ... ci_arttext


They compare the Headspace extraction with the SPME. The analytes are BTEX. Take a look. Tell me what you think...
They've reached with the headspace extraction and GC-MS a LQ for benzene of 0,0072 μg L-1. With a r^2 of 0,99 ! =)
The injected volume they use is 1mL. Double of mine. The split is 15:1.
The headspace volume is 10 mL. Same of mine. Vial of 20 mL, 10 mL water.

This one is very interesting too:

http://www.hpst.cz/sites/default/files/ ... -gc-ms.pdf

=)

[RogerBardsley]

The 80 deg C may be placing too much water into the GC/MS system which can interfer with the vinyl chloride. Application notes that I have done with our headsapce system (not a Combi Pal) typically works better at 65 deg C for static.

I am an application chemist for Teledyne Tekmar.

[seamoro]

i having the exact same problem with the vinyl chloride....did you figure out a solution?

[TiagoBele]

i having the exact same problem with the vinyl chloride....did you figure out a solution?

I just run out of septum!

I think that what RogerBardsley suggests it's a good way to get it resolved.
As soon the septum arrives i'll post my results.

Do you work with static headspace? Which equipment?

The 80 deg C may be placing too much water into the GC/MS system which can interfer with the vinyl chloride. Application notes that I have done with our headsapce system (not a Combi Pal) typically works better at 65 deg C for static.

I am an application chemist for Teledyne Tekmar.

Thank you Roger, i'll try it! Never think about this.

[TiagoBele]

i having the exact same problem with the vinyl chloride....did you figure out a solution?

Seamoro, it worked! Roger's sugestion works perfectly. With the concentrations of 1 ppb, 10 ppb, 40, 80 and 100 ppb. SIM mode.

I set the agitation velocity for 350 rpm (first was 500 rpm) and the temperature, like Roger said, to 65 degrees, (first was 80) . I've also increased the time to 19 minutes to reach a better equilibrium (first was 15 minutes at 500 rpm and 80 degrees).

Using a linear regression forcing to (0,0), all the coumpounds have a r^2 of 0,99! Vinyl Chloride have a r^2 of 0,95, which meet our requirements.

Guys, can i use quadratic regression on the Vinyl Chloride?
If i use it, the r^2 reaches 0,999 ! =)


Another two questions, and sorry for my lack of knowledge. First : the use of the internal standards and surrogates is necessary? because i'm using in my method the "target compound" for quantification. I don't see a reason for using them. Can someone explain?
And if it is necessary, how do i use it? I put them in my samples every time i'll do a analisys? I'll have to make a calibration curve for them too? A calibration curve with the other compounds?
Second: For example, we run analysis on different matrixes, like ground water and treated water. Do i have to make a calibration curve for all the matrixes?


=)


Thanks to all!

[Steve Reimer]

First : the use of the internal standards and surrogates is necessary? because i'm using in my method the "target compound" for quantification. I don't see a reason for using them. Can someone explain? If you are basing your method on 8260 for your 17025 acceditation, then yes.

And if it is necessary, how do i use it? I put them in my samples every time i'll do a analisys? I'll have to make a calibration curve for them too? A calibration curve with the other compounds? Read method 8260c and 8000c here: http://www.epa.gov/epawaste/hazard/testmethods/sw846/new_meth.htm

Second: For example, we run analysis on different matrixes, like ground water and treated water. Do i have to make a calibration curve for all the matrixes? If they are treated the same way then probably not but I am only guessing at your data requirements.

[lizardstone]

Another two questions, and sorry for my lack of knowledge. First : the use of the internal standards and surrogates is necessary? because i'm using in my method the "target compound" for quantification. I don't see a reason for using them. Can someone explain?
And if it is necessary, how do i use it? I put them in my samples every time i'll do a analisys? I'll have to make a calibration curve for them too? A calibration curve with the other compounds?
Second: For example, we run analysis on different matrixes, like ground water and treated water. Do i have to make a calibration curve for all the matrixes?


=)


Thanks to all!

If you don't prepare separate calibration curve, you should at least demonstrated that the surrogates and internal standards are within reaonsable amounr of variation in different matrices.

[memcadams40]

8260 HS can be tricky like all of them. Sorry I just caught this topic. The one issue I see off the top of my head is the low ppb serial dilutions in water. One thing to keep in mind is how much standard you are injecting into the water. For example - if you are injecting 1ul or 0.5 ul of a working standard into 500ml of water, you have more error possible. I found that always sticking to rcommended method volumes worked best. Try to stick with the dilution scheme in 8260. there is a reason the EPA validated and published the method procedure. Always try to make all your dilutions with the same Methanol standard containing all the analytes. Another thing is the range you are calibrating. Your spread may be too high to get a good R2 value. 1ppb - 200ppb and with multiple water dilutions--you'll have a hard time. VC is reactive and a dirty transfer line or GC inlet can pose a problem.