Chromatography Forum

Hi all! I'm having some issues and I was hoping you could provide some guidance!

I'm running an Agilent 7890/5975 GC/MSD with a Centurion autosampler/Evolution concentrator. I'm looking for help with the 8260B low level soil analysis. When running water and methanol extracted soil dilutions I have no problems. However, for the LL soils I am having problems with low internal standard recoveries (chlorobenzene-d5 specifically) in samples only. My theory is that it has something to do with the coelution of the deuterated and non-deuterated compounds in the calibration standards. For those of you who are not familiar, internal standard area recoveries need to be 50-200% of the area of the same internal standard in the continuing calibration standard. My thought is that when the calibration is analyzed (the internal standard mix is spiked during sampling by the autosampler) the coelution causes elevated areas for this compound. The same thing occurs with the CCV (and any spike). Then when the samples and method blank are analyzed the absence of the chlorobenzene from the calibration mix causes the recovery to be too low to meet acceptance criteria. This then often affects surrogate recovery as well.

Again, this occurance is strange to me because my system does just fine when using the sparge tube to purge waters or methanolic dilutions. So, is it a problem with the autosampler purge (since it purges in the vial itself)? The line from the purge vessel to the trap is longer than it would be for waters analysis but the desorb would be the same. Or is it the GC oven temp program? Or the MSD?

Any help would be greatly appreciated!

Thanks!

Just a note: I was actually hired to help fix the mess that was left by the previous manager of the volatiles department at my lab. I do not know what was done in the past or if they had the same issues that I'm having. My background is in 8020/21B which is PID/FID so I am new-ish to MS but am pretty fluent with P&T/GC.

Carrie

Do you also use 1,4-Dichlorobenzene-d4 as an internal standard or is Chlorobenzene-d5 your latest IS?

If you do use 1,4-DCB, does it do the same thing?

In my methods I am using Chlorobenzene-d5 with 82 as my quant mass and 117 as my qualifier mass which should not have interference from Chlorobenzene.

What are your IS concentrations and what is the concentration range of your calibration standards?

I actually use the same setup that you are using and I have a few posts here on some problems and observations I have made in the past with these types of problems. Let me know what your parameters are and I will see how they compare to mine. It is a common problem to have and some of us are still hashing out what exactly causes it.

Yes, I use 1,4-DCB as well. I was so focused on the chlorobenzene coelution issue that I didn't think about the fact that yes, the same thing happens to the DCB. My quant/qualifier masses for deuterated chlorobenzene are set the same as yours are.

My IS is 50 and the curve goes from 0.5 ppb to 200 ppb. I use the 40 ppb level as my CCV. My curves usually pass beautifully using a linear fit (r^2 >/= 0.995). ICVs are spot on as well. This is the same linear range that my waters curve uses.

Another possibility I was considering was that it uses a different pathway to get from purge vessel to trap but then if I am using the same parameters for the calibration levels and the samples then anything that is affected by this different pathway would be calibrated out.

I'm stumped. I will search your name and see what you have to say in your other posts. Thanks so much for your advice; any other tips you have would be more than welcome.

**I am using the Rtx-VMS 20mx0.18mmx1.00μm column by Restek and my oven program is that which is suggested (hold at 50 for 4, 18 deg/min to 100, 40 deg/min to 230, hold 3 min I believe).

Thanks again!

Since it is only with the soils then it could be some contamination in the transfer line from the soil purge needle to the trap. You can replace that using some 1/16" silcosteel tubing or what I have used is 0.53 silcosteel guard column or even the most inert is 0.53 fused silica guard column. You just have to be more careful with the fused silica for obvious reasons, but it does work well.

Another thing that seems to cause this problem is contamination in the Mort tube. Replace that and see if the linearity is better on the IS responses. I now only bake the Mort at 180C and it seems to remain linear for a longer time for both water and soil calibrations.

Thanks for the advice! I did a LOT of reading on the forum regarding the IS issue last night. My only question is this: I don't always do P&T maintenance before a curve (we rarely run low soils so when we do we just calibrate beforehand). A lot of posts were saying that the loss comes after dirty or suspect samples or even builds over time. However in my case, the curve runs, the icv then QC then samples. So essentially the cal levels and the samples are run under the exact same conditions. Is it possible that I'm seeing the char / adsorption effect that others were talking about?

Thanks again and I'm planning to try replacing/cleaning the MORT tomorrow. I do have a decent waters curve on the one instrument... If I see a difference in my later eluding compounds that may indicate that it's a potential fix!

VOAnalyst,

Hey; my name is Mike Sidler and I work with EST. There could be a quick fix for this problem. What revision of software is running on your Centurion? You can find that by going to the Options screen, then looking at the first tab System Info.

There was an issue (in older software) if you have a 5mL loop on your W/S, and you were adding 5mL of blank water to standards, but then adding 10mL of blank water to your samples. The IS was being spiked on the first of the 2 loop additions in sample runs (because to add 10 with a fixed volume loop, you need to add 5 twice). This was causing IS to recover low, so we revised the background processes to have the IS add on the 2nd loop addition. This gives 5mL of water to the vial before adding IS, so the IS not as absorbed by the soil samples in the vial.

If you want to contact me directly, my email address is msidler@estanalytical.com. That way, if you need it, I can send you the latest revision of Centurion software (version 2.3.01). Just let me know what version you were previously using for curiosity's sake.

I'd be happy to help more with this issue if the software update does not resolve it. Just let me know on here or via email.

Mike