PAH calibration problem

[erdno]

Dear all!

A have a 6890 agilent GC with 5973A MS.
I need to measure PAH with it but i have problem with calibrations.

We are using He as carrier gas and RTX-5SIL-MS ( 30*250*0,25)
The parameters are:
GC parameters:
Injector temp: 320°C, Splitless injection
Pressure: 8,73 PSI, Pulse pressure 36,3 PSI Pulse time: 1 min, Purge time: 1 min
Oven temp: 50°C for 1,5 min, then 20°C/min to 270, 5°C/min to 305°C, 40°C/min to 325°C and hold for 2 min
Column flow: 1,1 ml/min

MS parameters:
auotune
i have 10 SIM groups, MS source: 230°C and MS quad: 150 °C


I buy 2000 ug/ml PAH mix standard (contains 18 PAH + i make individually benz(e) pirene in acetone). I dilute them in 50:50% hexane: DCM solution. My calibration points are 25 ng, 100 ng, 400 ng, 1500 ng, 3000ng and 6000 ng/vial. I use naphthalene d8, acenaphthene d10, phenanthrene d10, chrysene d12 and perylene d12 internal standard for calibration ( concentration 400ng /vial and i buy mix for it too). I dilute my internal standards in acetone. I make everything in brown vials and storage in refrigerator.

I have problem with heavy compounds calibration. My curve fit is average of response factor. I can easily make 2-3% RSD for 13-14 compounds. My for last 5 PAH compounds i am unable to achive acceptable results (sometimes 50-60% RSD). At beginings a made my calibrations only in hexane, and i had the worst results with it. When i tried out 50:50% hexane: DCM, i could make good calibration for heavy compounds too ( all under 10% expect indeno(1,2,3cd)pyrene and dibenzo(a,h)anthracene 20-22%). It was about 2 week ago, now i wanted to again try it, i made 50 ng and 100 ng calibration points. For the first 14 compounds the recovery was perfect but the last 5 was horrible. The worst was dibenzo(a,h) anthracene i got 4-5 ng/vial instead of 50 ng, and 30 ng instead of 100 ng with my last calibration.

I cant understand my recovery of internal standards too. In my last calibration i got about 4 million area for napthtalene D8 for every calibration point and 3,2 million for my perylene d12. Yesterday I autoned again before i measured and for naphtalene i got about 3 million for every measure ( i tried it out with 8 different points) while for perylene 600 000- to 1,3 million aeras.

If it was volatile compounds, i could understand this thing. I tried out already so much thing if anybody got any clue what else to do i would be really happy!

Thanks for advance!
Erdno

[Peter Apps]

Hi Erdno

This is almost certainly inlet discrimination - try increasing the splitless time and the pulse time as a first step. Adsorption in the MS source is also known to be a problem with heavy PAHs, so increase the MS source temp.

There is lots in the archives - try a search.

Peter

[erdno]

Dear Peter!

I tried out this with 2,5 min purge time and 2,5 splitless time for a few compound it was a little better but i didnt help too much.

Erdno

[Amirav]

Hi Erdno
You problem is most likely due to ion source related peak tailing which depends on the ion source surface cleanliness and its history and which is particularly strong for the late eluters heavy PAHs.
You immediate remedy is to increase the ion source temperature to its max such as 300C.
You can read more on this topic and its ultimate remedy GC-MS with Cold EI at
http://blog.avivanalytical.com/2013/03/ ... hides.html
Amirav

[Peter Apps]

Dear Peter!

I tried out this with 2,5 min purge time and 2,5 splitless time for a few compound it was a little better but i didnt help too much.

Erdno

Purge time and splitless time are the same thing. Did you change the duration of the pressure pulse ?

What kind of inlet liner do you have and does it have packing in it ?

Also try Amirav's increase in source temp.

Peter

[erdno]

Dear Peter!

I am sorry wanted to write pulse time and spitless time 2,5 min. Yes i tried out to increase pressure pulse (from 30 to 50 PSI).

I am using at the moment single taper glass wool splitless liner.

I will try to increase ion source temp but i think my max is 250°C.

Erdno

[James_Ball]

Each autotune will give you slightly different settings. I would not run autotune each day or even weekly unless you recalibrate after each autotune. It is better to do a tune check to verify performance by injecting DFTPP and checking that the ratios of ions remain the same.

You may also need to be sure that your standards have warmed and been shaken or sonicated before using them as the heavy analytes can fall out of solution when chilled.

The heavy PAH compounds are know to be problematic. Agilent also recommends that you use the larger diameter draw out plate in the ion source to help combat the non-linear behavior. Agilent has also released a paper detailing some new technology they are using which inlets about 0.2ml/min of Hydrogen into the carrier stream as it enters the source to help reduce the memory effects and keep the source clean during analysis which improves the responses and linearity of the heaviest analytes.

[tangaloomaflyer]

when you say ng/vial is that equivalent to the on-column amount? i.e. when you inject the 25ng/vial standard do you end up with 25ng on-column?

[erdno]

Dear James!

I know autotune will give different results for each ion abundance but thats why i am using internal standards (5 compounds) so the abundance ratios should recejt this fluctiation, or i think it wrong?

I will try out sonicate the standards before i make the calibration points.

The new technology guess unable to reach with an "older" type MS (5973A), i am not sure if it is possible to exchange the ion soucre to an inert one too.

Dear Tangaloomaflyer!

When i say 25 ng/vial i mean 25 ng for each compounds in the vials (usually 25ng/ml). I inject 2 ul with autosampler so 50 pg on coloumn (with splitless).

Erdno

[AICMM]

Diff stack or turbo? Constant flow?

It has been a while since I have seen something similar to this but I have seen the pump choke at high temp constant flow and this leads to roll in response.

Best regards,

AICMM

[Hornet]

Hello, i have a similar method on my Agilent 5977 SQ and before this i did PAHs on a Varian Saturn 2000.
Your instrument conditions are good, ramp, splitless time and liner choice.
I work with similar conditions but i have the injector set at 300 °C (so no need of 320°C and certainly not an evaporation problem) and since we inject the same concentration (i work with 1ul at 50 ng/ml as lowest calibration point) i think your problem is in the MS.
My source stays at 280°C and transferline (which you didn't mention) is at 320 °C.

My advice is:

- Cut column MS end at least 20 cm, you probably have condensed heavy PAHs in there if you worked with low temperature transferline

- change liner if you did a lot of injection with it, i use restek sky single gooseneck with wool and they last more than agilent/other seller but after many injection especially with heavy matrix i start to have poor recovery especially on the heavy PAHs

- Take source at 280 °C for sometime (if your instrument can reach this temperature) and try to reinject your standard.

Try to stay at 1 ul injection if your instrument allows you to and to avoid acetone at high % in your final solvent, in the past i had precipitation problems on perylene-d12 when working on the varian and injecting 1:1 DCM/acetone extracts, but honestly i don't think it's your problem.

[erdno]

Dear all!

I tried out my calibration points on an old GC-FID ( HP 5890). First i measured it by my GC/MS and it wasnt so good. Then i measured it with GC-FID on a DB1 column. My carrier gas was H2 with constans pressure. I used perilene D12 for internal standard for all compounds since i could separate it easily. I could do really nice calibration for all heavy compound. (however it wasnt so easy because of low concentrations).

Dear James!

I tried to sonicate the standards before using it, i think it helped a lot (since with GC-FID it is good and with GC-MS it is better).

I dont really know what u mean by that: "gilent also recommends that you use the larger diameter draw out plate in the ion source to help combat the non-linear behavior"

Can u explain it a litte pls?


Dear AICMN!

This type has 2 stage pumps with a turbomolecular pump. I am using constans flow.

Dear Hornet!

My transfer line at 315°Cat the moment, and i think my max ion source temp is 250°C. I tried it out and i got higher responses in heavy compounds (compare with 230°C).

I have acetone only in my internal standard and in my individually benz(e)pirene solution. It is so funny i can make great calibration for benz(e)pirene but horrible for benz(a)pirene... The other compounds in 50:50 hexane:DCM. Perhaps i will try out 100% DCM perhaps it helps too.

Thanks all!

Erdno

[Steve Reimer]

The draw out plate choices are listed on:
http://www.chem.agilent.com/Library/sel ... es_QRG.pdf

The 6 mm plate is recommended for extended linear range.

[Hornet]

I have acetone only in my internal standard and in my individually benz(e)pirene solution. It is so funny i can make great calibration for benz(e)pirene but horrible for benz(a)pirene... The other compounds in 50:50 hexane:DCM. Perhaps i will try out 100% DCM perhaps it helps too.

Thanks all!

Erdno

No your solvent is good as it is. It seems you'll have hard times to make it work well for the heavies with that source temp.

Davide

[James_Ball]

Dear James!

I know autotune will give different results for each ion abundance but thats why i am using internal standards (5 compounds) so the abundance ratios should recejt this fluctiation, or i think it wrong?

I will try out sonicate the standards before i make the calibration points.

The new technology guess unable to reach with an "older" type MS (5973A), i am not sure if it is possible to exchange the ion soucre to an inert one too.

Dear Tangaloomaflyer!

When i say 25 ng/vial i mean 25 ng for each compounds in the vials (usually 25ng/ml). I inject 2 ul with autosampler so 50 pg on coloumn (with splitless).

Erdno

At 25ng/ml (25ug/l or 25ppb) that is really pushing the limits for a 5973 in full scan mode for most analytes. Our normal calibration had 5ug/ml as the low calibrator. I have been able to do 10ng/ml for PAH compounds on my Agilent 7000 using SIM with good results but is difficult to make a linear calibration down to that level running full scan even on that instrument. You may want to try using SIM to see if you can get better results at those low levels.

Our analysis consists of running standards in low ppm and extracting the samples to give a 1000x concentration factor so that we can quantify in the ppb levels.

Another thing to think about, injecting 2ul will not give a 2x sensitivity increase over injecting 1ul, so maybe it would be better to try 1ul injections so that you might get better evaporation of solvent and less solvent load on the column for better results.